What PSRU Can we Really Trust???

Hi;

Well, I've been meaning to ask this for a while. I find it frustrating that a very critical part of many homebuilts is also perhaps the least understood, regulated, or reliable. OK, if re-drive manufacture were truly regulated by a governing body, then costs would be even more through the roof than they are right now for a gear box or a belt drive. But, this has begged the question, at least in my mind, "What manufacturer of PSRU's can be trusted?"

Those who might want to build a fighter replica have some choices, but for engines, the V6 and V8 power plants seem to be likely candidates, based on the shape of the engines. It seems wrong to stick an opposed 4 onto a scale Spitfire or Mustang, and even though I'm sure the Isaac's Spitfire is a wonderful piece of work, I just don't like the look of it. After all, it is supposed to be a replica fighter.

So if we use a V-engine, who's re-drive do we trust? The makers of gear boxes seem to come and go rather quickly. Those who are still making these things may not be real experts about that enemy, torsional vibration (even if they claim that Hyvo chains or belts damp out torsional vibrations because they have a slack side).

Vesta's Jason Day uses Hyvo chains, and states that vibrations fed back through the system to the engine are damped. If you talk to Jack Kane of EPI (the only guy who REALLY seems to make an effort to address the issue of a variety of vibration modes), he'll tell you that this assumption is wrong and potentially dangerous. Having said that, others may know if EPI's products, notably its Mark 9 gear box has ever suffered failures. (I really don't know).

Dan Hawken of Alberta designed his re-drive for the Titan-51 with a 3" Goodyear Hawk Kevlar belt and cogged pulleys. Titan has offered the Suzuki V6 2.7 litre as an alternative to the Rotax 912 for a few years now, and seem to have had a good run with this setup. But, they've just experienced their first failure, a crankshaft that, upon 3rd party inspection, was found to be a victim of torsional vibration. And this was a forged crankshaft (the strongest type, is it not?).

Now Titan is adopting the New Zealand gearbox. This maker claims that some units in Australia have over 1000 hours of sevice on them. When I write to the company about costs, availability, and any failures that may have happened throughout their history, I don't get any reply. This bothers me a bit, and the information trail ends there. Maybe they're just jammed busy.

Can we all name a few of these PSRU makers that have come and gone in the last decade or so? Probably.

So here's the bottom line: if we want to use a V engine, from 150 to 400 HP, who's reduction drive would YOU use, and why?

For a guy who's not an ace at torsional vibration (and never will be) this is really a case of trust, hopefully built on some manufacturer's documented, reliable track record.

I'm thrilled about the idea of using auto engines, but very concerned about reliability.

Thanks, Tom.

This has been a continual frustration for me also and unfortunately, my research into the subject matter has not been any more fruitful than anybody else's.

So, the drives I like include the one for the Mistral rotary and the one manufactured by Powersport. True, these are not for V-8 application but they are representative of drives properly designed not only for the drivetrain loads but also for loads that are imposed onto the structure by the prop.

The only drive whose design I liked for the V-8 (outside of the Geshwendler) was one developed by a company from Canada - it was called a Marcotte drive (Marcotte PSRU - geared aircraft engine reduction drives and SDS EM-2: Aircraft). But unfortunately I had an unpleasant experience with them (deposit but nothing delivered) so I certainly can't vouch for their business practice.

The only other drive that caught my eye is one for airboat use (: ROTATOR REDUCTION) so I don't know if anyone has ever installed it on an airframe but the design looks to be robust and quite sound. Given the beating and rough service the airboats normally see, if this drive has a good reputation, it should serve an airplane application quite well.

It's kind of depressing to think that every single propeller reduction drive ever written about has had bad things said about it. Whether by armchair experts or those who really know their stuff, but has anyone ever read of a drive system that was universally praised? Except by those selling them?

Personally I am putting my money with the above mentioned Rotator unit, not only because it seems sound, but also because it has a halfway reasonable price tag.

This question is slightly askew the topic and asked from a position of total ignorance, but would it be possible to make a reduction drive that is both gear driven and belt driven so that the load could be carried by either or both? Would you gain any benefit from it (i.e. - redundancy or harmonic suppression)? Mechanically, I can see where it might be tricky to get the gear ration exactly synced with the belt ratio (at least for a cogged belt).

Bruce

PTAirco;

I've never heard of the Rotator unit, so I can't comment. Maybe I should look it up and have a peek, but I have a thought about the price.

It really seems in the PSRU business that you don't necessarily get what you pay for (like the old adage claims you should). I mean, sure, you might get a lot of stuff. A PSRU is more than a couple of cogged pulleys, a belt and a housing. Just for fun, have a look at Titan's PSRU package. Geez, that's a ton of stuff for $16,000. Laid out on the shop floor, it looks like an archeological dig. But does that mean it's good?

Similarly, does that mean that a unit is good if the price is reasonable? Even the unreasonably priced ones break. I don't have any facts and figures, I just recall reading about several different PSRU failures, and related engine failures, due to some shaft twisting itself to death. Scary stuff, at full power, on take off, 100 yards into the neighbour's corn field.

So in fairness to the Rotator unit, I should have a cursory look, even though it won't tell me the full story.

Tom.

And of course, I forgot the attachment.

Tom.

The knottiest problem in Engine-PSRU design is torsional resonance, and torsional resonance comes in an extra mean version with the Wankel rotary. The Powersport PSRU that Everitt Hatch developed beat that problem by putting together a preloaded gear set using 8 pitch gears. The combination of zero lash and big gears (oversized for the torque involved) did it. It gave a lot of stiffness, shoving the primary vibration mode between the prop inertia and crank inertia right off the high end of the rev scale.

Funny thing, oversized gears and lots of stiffness is what worked in the Rolls Royce and Allison engines too.

If'n I were to bet on a system with minumal issues, it would be one with big gears and very low to zero lash in the gears. So, do your check with the makers and ask about lash. Look for stiff strong systems, because flexible systems can bring the primary vibration mode down into the operating range...

And using a Detroit V8, I would stick a Rattler on the "front" end of the crank too. The torsional pendulum works like magic for sucking off the big vibrations.

Now I had never heard of these airboat drives, but they look serious, and if they have a good reputation, so be it. Another one that looks serious is The only spur gear PSRU with* ce. I looked at his redrive at OSH. Sturdy, well built, and even cold, you could only barely feel the lash in his redrive. Oh, and he includes a OWC in the system to reduce any feedback of resonant vibration. And his redrives go to the shows on airplanes that are flying regularly....

Billski

In regards to the "Rotator" PSRU, it looks like it has two spur gears, so the output rotation is reverse (CW - Looking from the flywheel end). I was hoping this was a chain drive (CCW). The price looks good; The design looks good too.
Anyone know more details on it?
Jan

I've been thinking about what makes a PSRU a good one. I'm starting to think that it's really an issue of matching the whole system together. I'll explain what I'm getting at:

We could look at this as 3 parts, the engine, the PSRU, and the propeller. I suspect that if you stick a good PSRU onto the wrong engine, things will go bad. I have read about a few crankshaft failures in my surfing, but this is an engine part, not a re-drive part. So the re-drive survived, the engine didn't.

It doesn't always work that way; sometimes PSRU's fail. I've read about those too. But you did mention the Rolls engines, and in the case of the Merlin 60 series, there is actually quite a bit of backlash between the output shaft and the pinion gear which mounts to the front of the crankshaft. It's not huge and sloppy, and you can definitely feel the big gear going "ping" as it smacks against the pinion, when you twist it back and forth. So, I'm not convinced backlash plays a critical role in this, as long as it's not stupidly big.

What I'm starting to think, is that crankshafts are not big enough on auto engines to be used in aircraft. In truth, I don't know how much bigger the throws and journals might be on aircraft engines (if any) but when is the last time you heard of a Continental 0-200 or Lycoming O-235 crank failure? Then again, given my complete non-understanding of torsional vibration, maybe it isn't a crankshaft size issue, maybe it's much more than that.

So why does a Merlin survive? The weakest link on Merlins, especially for the racers, seems to be the connecting rods. And I'm sure they are running lots of different propeller designs/weights on the shafts, which would change the vibration dynamics, I would think. However, you can't argue about the beefiness of Merlin reduction gears, because they are beefy.

The only issue I recall reading about Merlins and front end damage, was some of the P-51 blades, the newer hollow ones, sometimes broke, so they stayed with the solid metal ones after that.

I wish I had a better understanding of the problems of PSRU design, and how they match with engines and props, but this is one area where I will never gain a full understanding. Too complex for me. And unfortunately, that ultimately means trusting a PSRU manufacturer if I use an auto engine.

I talked to Gale Derosier yesterday, who built a Midget Mustang II with a Chevy 4.3 V6 Vortec, and a "Belted Air Power" redrive. 200 hours, no problems so far, but something that really surprised me, on the BAP website, they state that they use cast iron crankshafts (stock for the Vortec) and don't switch to a forged. I wouldn't have expected this. No reply yet from BAP to my enquiries about any failures they've logged in their years of existence. I hope owner, Jess Meyers writes back.

Tom.

For those who want to learn more about vibration, as it relates to PSRU's, I recommend Jack Kane's site, "EPI." Although I cannot attest to the success of EPI's re-drives, I am heartened to see that somebody wants to tackle the issue of vibration, in all its complex forms.

I also love the bashing he gives one maker of PSRU's who claim that a one-way clutch eliviates the concerns of torsional vibrations transmitted from engine to propellor (and backwards).

I hope this link works, and I did take time to read the whole page, although there is a ton of other info on the site.

PSRU Fraud: Sprag Clutches to Eliminate Torsional Vibration, by EPI Inc.

Tom.

Originally Posted by Tom Kay

...Titan has offered the Suzuki V6 2.7 litre as an alternative to the Rotax 912 for a few years now, and seem to have had a good run with this setup. But, they've just experienced their first failure, a crankshaft ...

A completely uninformed opinion, but can it also be that we're setting unreasonably high standards for "success" and "failure"?

I completely understand the importance of avoiding powerplant failure, but if a single failure "in a few years" is cause for rejecting a design, I think we may have gone too far. Lycomings and Continentals (without PSRU) fail occasionally, and yet we don't throw them out as "dangerous" or "unsafely unreliable". Everything breaks. You can't build a 100% reliable mechanism, especially in the heat and vibration environment of an airplane powertrain, and with the stringent weight requirements we have.

In this particular case you mention above, well, sure. Sounds as if the design can be improved. But anything can be improved. Just because it's possible to improve something doesn't mean it's a failure. I'm not suggesting that we accept sub-standard designs, but it's also possible to set a standard much too high. Nothing will meet it. That seems to be the case here, IMHO. One failure? In several years? That's pretty good from where I sit.

I'm just concerned with the overall suggestion that if a PSRU design fails once, even after a significant history of service, it should be rejected. It's no wonder there are few companies able to survive this market if that's the case. One mechanical failure and they're doomed commercially.

Topaz,

These are goods points you've made. Everything breaks. I guess one of my number of concerns is false or absurd claims by a company if they have no basis in fact, and the designer is making unproven assumptions. Do Hyvo chains or belts really attenuate torsional vibration, just because they have a slack side? Or is that utter nonsense?

It would help if every maker of re-drives kept good stats on their products and made them public, or at least available to those considering a purchase. If, for example, I saw a few failures in a design, then evidence of substantial improvement of design, followed by years of successful flying with tons of breakage-free hours, I'd probably be willing to overlook the developmental years.

If, however, I saw a design where the chain, or the crankshaft of the engine breaks surprisingly often, or even a wide scattering of problems in a single design, well, I'd like to know this before plunking down thousands of dollars (and my life).

I still think it comes back to this: despite any due diligence on the part of the homebuilder, it comes down to trust. And it's like being a Creationist or a Darwinist, do you have faith, or do you need proof?

However, your point about not scrapping a whole program because of a single failure is a relavent one. (Sounds like a line from Dr. Strangelove).

Tom.

Originally Posted by Tom Kay

I talked to Gale Derosier yesterday, who built a Midget Mustang II with a Chevy 4.3 V6 Vortec, and a "Belted Air Power" redrive. 200 hours, no problems so far, but something that really surprised me, on the BAP website, they state that they use cast iron crankshafts (stock for the Vortec) and don't switch to a forged. I wouldn't have expected this. No reply yet from BAP to my enquiries about any failures they've logged in their years of existence. I hope owner, Jess Meyers writes back.

Tom.

I remember digging through some books on automotive design and reading that nodular cast iron crankshaft have a useful capcacity for absorbing vibration through hysteresis. The absorbed energy of these vibrations is given off as heat. I am not sure just how much they are capable of absorbing, but I would not automatically assume that a good quality cast iron crank is a non-starter in an auto conversion. I am sure the loads on the crank of a souped-up Vortec engine in a Hot Rod are quite staggering and they seem to survive well, both in V8 and V6 versions.

I just read a lot of stuff on EPI's website. They know the topic of vibration well enough to write about it well. When I read their commentary about the one-way clutches, I was a little surprised. But then, the opinions of people you talk to depend on what they are selling. Me, I am not selling anything...

First my credentials - Degreed engineer, 18 years on automatic transmission design, the last 9 years as a management grade engineer, and vibration control was a big piece of my job.

Conventional automatic transmissions usually have several OWC's in each one. Their mechanism is simple and robust, they are rotationally fairly stiff and very fast acting. The suppliers may not test them faster than 40 Hz, but there is nothing about them that inherently makes them fail to release if the torques on them cycle faster than some specific speed.

More important to their ability to keep vibrations from amplifying is how far they travel from loaded to unloaded in the release direction. A OWC is spring loaded to engage, and as it picks up torque, the inner and outer races deform slightly and rotate slightly relative to each other. To unload the OWC when a reversing torque pulse comes through, the OWC must move back through that deflection it went through to load up.

Now if this relative rotation is larger than the oscilatory travel of the vibration, yeah, it will not release and slide. I did some math. The nominal firing acceleration of a gasoline V8 engine at full throttle is about 2500 radians/s/s. At 1000 rpm, the oscilatory travel will be 1.63 degrees, at 2000 rpm, travel is 0.41 degrees, at 3000 rpm, it is .18 degrees, and at 4000 rpm, it is 0.10 degrees. These numbers sound pretty small, but the elastic travels of OWC's are pretty small too, something less than 0.5 degrees. They will transition from releasing during oscillations at low speed to not moving enough to release at higher rpm with just the V8 firing pulses for motivation. If the OWC's are bigger and stiffer than the ones I worked with (they have to be), they will need even less travel to release.... Additionally, we are not worried about the basic firing pulse, we are worried about when the pulse is amplified, and the travels are trying to be double or triple... Well, the OWC well then release on the rebound to higher rpm. Yeah, they might not cover the whole engine speed range, but they will help with much of it.

What happens with the flat four Subie? All of the oscillatory vibrations are four times as high, and the OWC will release through an even larger part of the engine's rev range, and certainly can contribute to preventing amplification...

So, EPI's claims that the OWC is not doing anything is an exageration.

EPI's comments about a windmilling prop being able to overspeed in a fast descent, if you have a fixed pitch prop, its max safe rpm may set your max dive speed. Even with a constant speed prop, this may come into setting operational limits. Could be an issue...

Last is the issue of it being less draggy on approach and inital decel after touchdown. Yep. It is there. Maybe you should consider a dive brake on the belly...

Which would I pick? So far, the one that has the most sold without getting bad press. But it should have beefy gears, very little lash, a stiff case and beefy shafts, and stick the Rattler on the other end of the engine if you can get it... And I would not let the presence or absence of a OWC make up my mind all by itself.

Billski

Do you not think the Re-Drive cops a lot of bad flack when its really the combination of prop-redrive-engine that has not or cannot be matched. A/C engines have a 'locked down ' list of components, when you get into the Auto Engines with Re-drives such as the Stewarts etc, its highly unlikely that any two combos are the same, one guy fits a steel stroker crank, uses 'Brand X rods & pistons, next guy uses a cast stroker crank, Brand Y rods & pistons. Neither crank is going to have the same torsional vibration characteristics or even the same bob weight factor when balanced, Sooo the poor guy building the PRSU has no set guidelines to even consider having to factor into the unit. I spoke with Fred Geschwender about this very scenario many years ago & he agreed that in his initial endeavours that they were swayed by the opinions of many so-called experts that they needed to 'do this' & 'do that', but in the end they went back to standard Ford Engines & Factory Balance--- that gave them the least problems.