What PSRU Can we Really Trust???

Jake;

I think I have read about at least one geared PSRU that had a chip detector, but even with this equipment, isn't it a case of a shaft giving up the ghost without apparent pre-damage? I mean, in the case of a crankshaft, or input shaft, that's twisted back and forth to death, do you get chips? Or just sudden breakage?

Yes, Billski has put some commendable effort into helping us understand the complexities of the full "system" as has that EPI website (which I suggest as reading material for anyone).

Tom.

Jake,

Chip detectors are COTS, and should be more widely used in light airplanes. Remember that they detect metal pieces in the oil. If a bearing or a gear has had enough sub-surface damage (occurs as one surface rolls across another many many times), you begin to free pieces, and that is what the chip detector is able to see. Hopefully, it spots the chips soon enough to be able to abort the mission and safely conclude the flight.

For the topic of PSRU's, they can be useful - We have bearings and gears that could tear up, and the warning could prevent a dead stick landing, but will only give us warnings on the failure mode mentioned. They give virtually no warning on others. Those have to be precluded by design. If a shaft or case is being fatigued, generally single cracks will emanate from a tiny flaw in the surface of a part, and it will grow with successive load cycles until the remaining area is reduced enough to cause complete failure of the element. Generally no chips are released by this process until the end. You could detect the cracks by disassembly and any number of non destructive methods, but you generally will not see them with the Mk I eyeball or the chip detector. Magnaflux, dye penetrant, etc are very useful for catching cracks in parts at rebuild time if the parts in question have long but not infinite life.

When we are talking about PSRU's managing torsional vibration inadequately, we are talking something that should easily last way over 2000 hours that instead is lasting minutes because torsional vibration and feedback is increasing the loading by double or triple or even more over what was expected.

So, useful? Yes. The careful operator, could use them for the same issues that they are in your Scout, but only after the vibration management issue is properly addressed. The chip detector would not be very useful for detecting flawed vibration management, as the detector warning will not sound until just about the time the powerplant goes into death throes...

Bill

Originally Posted by wsimpso1

...Topaz: In this context, soft pertains to spring rate not material condition.

Designing it to live and to give the right spring rate may conflict, but they are not mutually exclusive.

Deflection angle of a torsion bar element is T*L/(G*J) where T is torque, L is length, G is torsional modulus of material, and J is torsional section modulus. Spring rate is thus G*J/L. Make it long (big L) and slender (small J) and it will have a low spring rate. Build it with low G and it will have a lower spring rate too. Oh, and J goes with diameter to 4th power... Now shear stress in a torsion bar element is T*r/J. So the design task is to obtain the right spring rate at a low enough stress. It can be and is done. One thing to remember is that the light way to make a stiff shaft is large diameter and thin wall, while the way to make a soft torsion element is to make it small diameter and long. A lot of attention has to be paid to ends and stress concentrations and heat treat. And then you have one springy shaft. Worked great for isolating pumps and generators and the like from crankshafts....

Ah, okay. I see what you're getting at. I'd seen successful applications with tail-rotor drives and such, so I figured there must be a flaw in my thinking somewhere. I was missing that the ability exists to tailor the spring rate and still keep a low stress level. Thanks.

Bill;

For those of us who aren't up on our acronyms, what's a "COTS?"

Tom.

Originally Posted by Tom Kay

Bill;

For those of us who aren't up on our acronyms, what's a "COTS?"

Tom.

I think he means "Commercial Off The Shelf".

Originally Posted by Topaz

Sounds as if the only sensible way to go is to sell engine/PSRU combinations as a unit - no individual sales of the PSRU. Limits your market on one hand, but your liability on the other.

This is why Jason at Vista will only sell combos...

Bill;

I've read your comments a couple of times which gives me reason to believe that you can offer some good advice. The problem is I'm not a mechanical engineer (adhesives and bonding aluminum, now I can do that stuff all day long until your eyes glaze over), and I'm not sure of all the terms you're using (one of my educational shortcomings).

For example, you refer to an 'OWC,' I must apologize but I don't know what that is, nor, what a 'Rattler' is, which both seem to be useful in curbing torsional resonance problems (if sized properly).

Should I want to use a Subaru engine as an alternative power plant and the same PSRU that Ram Engines uses (also available directly from New Zealand), what would you recommend as a 'belts and suspenders' approach to the technically challenged (specifically me)?

This thread has been most informative, and your comments lead the pack.

Thanks!

Fly safe,

MalcolmW

Yeah, COTS is Commercial Off The Shelf, meaning you can buy it now and plug it in...

OWC is one-way-clutch. There are two common types, sprag and roller, and they allow torque to be carried in one sense, but overrun if you try to make them carry torque of the opposite sense. Used in a lot of machines, including at least two in every automatic transmission I know of... Hit Wikipedia for way more...

Rattler is a commercially available tuned pendulum vibration damper. Right now, they are available for most Detroit V8 and V6 engines. They work pretty well, but they put out a short but distinct rattle at engine shut down or start up. The company decided to just make a point of it by naming it that... Wiki on "pendulum damper" for lots more.

Read my writing close. I don't know what WILL work for any application. I have been warning of the pitfalls that will give you no warning, but could suddenly give you a powerless airplane... Without actually putting torsional vibration equipment (a $60-100k set of equipment that I no longer have access too) on a system, I can not tell you if there are issues with any particular setup.

If other folks are using a particular setup successfully (several powerplants with a thousand hours each) I would say it is a good bet. But changes that seem insignificant cause vibe problems in cars, trucks, trains, motorcycles, airplanes, ships, machine tools, commerical powerplants, etc. It is real...

Billski

Since we have PSRU discussion already, I'm wondering, does anyone have any info about Thunder Engine TE495-TC700 (attachment1) and Bonner Sapphire V6 (attachment2). Unfortunately I have only those two pictures, no more info.
Seb

So what have we learned? Would it be fair to say that the average aircraft enthusiast doesn't really understand torsional vibration, or am I the only one? Taking this further, could we say that the average PSRU maker doesn't either? I mean, are claims about cogged belts, hy-vo chains and one-way-clutches absorbing/attenuating/defeating this type of vibration fair or nonsense?

If the goal still remains to have a power plant based on a V auto engine, how do we go about this? And I know you could ask for a range of engines, the 150 HP class, the 250+ class, the 400 HP class, it depends on what you're building.

I think I can accept that the entire system must be matched to really be reliable, but again, who on the open market offers this, and has the years of successful service, plus the vibration survey to back it up? I can accept the concept of not scrapping a PSRU program just because of a few "failures" as was mentioned earlier, but I guess I did mean fwf failures of any kind, either prop breakage due to vibration, crankshaft breakage or PSRU failure. I guess it's not fair to simply blame the reduction unit.

All of the replies are indeed appreciated, but at this point, I'm still not sure where to turn for a 150 HP fwf package to power a 3/4 entry-level performance fighter.

Still eager to see further discussion, plus ultimately, a magic solution, of course.

Cheers all, Tom.

Originally Posted by Tom Kay

So what have we learned? Would it be fair to say that the average aircraft enthusiast doesn't really understand torsional vibration, or am I the only one? Taking this further, could we say that the average PSRU maker doesn't either? I mean, are claims about cogged belts, hy-vo chains and one-way-clutches absorbing/attenuating/defeating this type of vibration fair or nonsense?

Stay away from Hyvo chains and belts, no good, must use gears to have a good prsu.

The gearing needs to use needle bearings, with oil chutes coming from the center spline on the bearing shafts.

IMO Its not torsional vibration as much as it is torque reversals that wreak havoc on the whole gearbox. A tight tolerance can help but will cause fatigue due to the tooth pressure, the one way to fix this is to have thicker gears and teeth along with a curved(arched) tooth edge. The contact are must also be larget to dissapeate the psi. The mental used for the gears must be between 350 and 450. I would also think that a shot peened or nitrated gear set would hold up best.

torsional vibrations become a problem as the engine becomes unbalanced, this is because it is enharmonic in nature, Likewise a inline engine gets smoother as its RPM rises.

AS a V engine increases its RPM it does have torsional vibration issues, and so yes torsional vibration can be an issue at high rpm situations.

The most important thing is that V-6's are enharmonic, and are heavily influenced by torque reversals. V-8's and V-12's, (v-10's with odd firing orders) are all inherently balanced and can use a cross plane crankshaft.

However as everything gets larger we must remember that the reversals will get larger, and as such so will the stress on the engine.

I think I can accept that the entire system must be matched to really be reliable, but again, who on the open market offers this, and has the years of successful service, plus the vibration survey to back it up? I can accept the concept of not scrapping a PSRU program just because of a few "failures" as was mentioned earlier, but I guess I did mean fwf failures of any kind, either prop breakage due to vibration, crankshaft breakage or PSRU failure. I guess it's not fair to simply blame the reduction unit.

Crankshafts break because of torque reversals that exist on the crank's order, as long as you don't exceed the engines design specs you should be fine. you can minimize and or eliminate TQR by adding balancing shafts.

Reductions units that are designed properly will not break, rather wise it is the person who doesn't research into it as much, the best way to keep a prsu attached is to have it molded into the block, and then have the crank end geared, have it a monolithic casting. When you start adding moving parts bolts, chains etic you add more stress to any PRSU, and in term stress the system,. that is the problem with HyVo chains, as they wear they loosen up, and then the torque reversals eventually cause them to stretch so far that they fail.

All of the replies are indeed appreciated, but at this point, I'm still not sure where to turn for a 150 HP fwf package to power a 3/4 entry-level performance fighter.

Still eager to see further discussion, plus ultimately, a magic solution, of course.

Cheers all, Tom.

(check last ppg for my suggestions for a motor)

IMO the best is a 2 stroke outboard engine and then design a PRSU around the bottom lug attachment (where the shaft goes into the engine. Running gear reductions must be kept appropriate, you must have a idler gear that is proportional to the RPM. meaning the engines gear should be large enough to allow for low speed operation, however not to large that it reduced your overall RPM band.

My solution to the gearbox issues with my engine designs was to incorporate the back of the gearbox into the bottom block casting. And then to have the shafts cast in and then milled out. the spindle shafts that the gears would sit into were full of needle bearings, the bearings ran the length of the shaft with multiple oil ports where the oil could escape. the gears were big however a good 2.5 inches thick. My bottom gear was larger then the original Merlins pinoin gear.

For example here is a Merlin gearbox assembled
http://www.enginehistory.org/ModelEngines/Hares/Merlin%20XX/1b.jpg


This picture shows the gearbox dissambled
http://www.enginehistory.org/ModelEn...in%20XX/1a.jpg

Notice the use of roller bearings, which help with torsional stability, however the needle bearing i believe will increase high speed stability, especially with the new oil porting system i designed.

The oil is inserted through the center of the spindle shaft, around the spindle shaft there are slices and holes where the oil is pushed into the bearing jacket VIA a high pressure oil pump. The oil is then pushed into the bearing where it finds its way out of the outer spindle shaft, the outer spindle shaft is what the upper idler gear is attached to.


Tom check this out I have read this set of papers many times over, and i feel you might find this web page useful
Gearing

It talks about anything and everything about Merlin PRSU's.

IMO an updated version of this gearbox would do you the best. Along with a Evinrude E-tec 2 stroke EFI outboard engine. They will produce 300HP at 5,000 RPM, with a simple injector and fuel pump change you could pump out another 30 or 40 hp if not more depending on your poison, and at 500lbs its 200 lbs lighter then the 4cycle motors you can find.

Its quite possible to pump out 150 HP at 3,000 rpm. The engine is super light, and very powerful and IMO would give you the best HP:LBs ratio you could get out of any engine package.

Tuningout a 2 cycle is very cheap, easy and its reliable, with the least moving parts of any engine design.

Any drive has the potential for being good as long as it's properly designed for the loads, service environment and service conditions. Making blanket statements as to what's good or what isn't, especially when those statements have no basis in fact, is a disservice to potentially good product lines, even though in this case we really don't have a lot most of us would be willing to bank on too much.

Originally Posted by pepsi71ocean

Stay away from Hyvo chains and belts, no good, must use gears to have a good prsu.

Properly selected silent chain drives can work and work well. The fine pitch variants have very high surface speed ratings, making them suitable fro application even to rotary engines and other high rpm products. They have multiple contact geometries which means that sprocket tooth stresses are low and thus the potential for catastrophic failure can be significantly lower than what might be envisioned for a gear drive.

True, I am not a fan of smooth or cog type belts for our application but that doesn't mean that one could not be properly designed. Some years back I worked on a large prop drive for a hydrofoil (two C-130 shrouded props up top) where we initially were going to use the HTD cog belts. BTW, we did not go that route, not because the belts wouldn't have worked but because at that size and power level they were simply way too loud.

Originally Posted by pepsi71ocean

IMO Its not torsional vibration as much as it is torque reversals that wreak havoc on the whole gearbox.

What do you think torsional vibration is? It is just that - torque reversals created by the spring characteristics of the engine/prop system which when operating near its natural frequency (or some harmonic thereof) create very large torque pulses that can rather quickly exceed the design specs of even the most robust system.

Originally Posted by pepsi71ocean

vibrations become a problem as the engine becomes unbalanced, this is because it is enharmonic in nature, Likewise a inline engine gets smoother as its RPM rises.

This is an internal issue (engine only) and a consideration if you're designing your own engine. Most folks are not that ambitious. We just want to use a good off the shelf product, coupled with a reliable reduction.

Torsional vibration has nothing to do with internal balance - in regards to our discussion, it is only that characteristic which is defined by the behavior of the engine/drive/prop system and its response to pulse input and spring based (prop and drive) feedback.

Originally Posted by pepsi71ocean

.....the best way to keep a prsu attached is to have it molded into the block, and then have the crank end geared, have it a monolithic casting.

I strongly agree but that represents quite a sizable investment. By the time you're done, and given the limited market, I'd guess that the product would pretty much price itself out of most folks' ability to buy.

Originally Posted by pepsi71ocean

the best is a 2 stroke outboard engine and then design a PRSU around the bottom lug attachment (where the shaft goes into the engine. Running gear reductions must be kept appropriate, you must have a idler gear that is proportional to the RPM. meaning the engines gear should be large enough to allow for low speed operation, however not to large that it reduced your overall RPM band.

We've talked about marine engines before and I agree, it's a great idea, especially given the technology of some of the new engines. The problem though is simply that the cost of these power-heads is quite astronomical - might as well buy a more conventional aircraft engine. By the time you add all the peripheral stuff you'll have another gold-plated paperweight since no-one will be able to afford it.

We've discussed belt, chain, and gear redrives, but what about hydraulic, magnetic (as mentioned in an earlier post), or a mechanical-electric-mechanical system (like they use on trains where the engine drives a generator and the power drives electric motors at the wheels). I know each has it's own drawbacks and weight penalties, but are any of them in the ballpark to be viable?

Originally Posted by bmcj

We've discussed belt, chain, and gear redrives, but what about hydraulic, magnetic (as mentioned in an earlier post), or a mechanical-electric-mechanical system (like they use on trains where the engine drives a generator and the power drives electric motors at the wheels). I know each has it's own drawbacks and weight penalties, but are any of them in the ballpark to be viable?

All of those are really heavy. They're fine on locomotives and mine trucks and other heavy equipment. We even find hydrostatic drives on lawn tractors; got an older Bolens here that has it, but that system is awesomely heavy compared to the cheaper belts and gears used on other tractors. This tractor easily weighs twice as much as a Sears tractor of the same HP. An airplane using hydrostatic drive would be a real slug.

And all that stuff adds more failure points and has enormous efficiency losses compared to a PSRU or direct drive. If it made practical sense someone would have built a workable one by now. This topic has been bouncing around the homebuilt community since I joined it in 1972 but I don't know of anyone flying such an animal.

Curiously, we have an old (1966) Massey-Ferguson backhoe here that has, believe it or not, Cessna hydraulics on it. The valves and other bits are all made of cast iron and aluminum and have the old Cessna logo on them. What reason Cessna had for being in that industry is beyond me, unless they had tried the hydrostatic drive on some airplane and had to recoup huge financial losses somehow.

If a guy wants to fly in his lifetime without spending everything he makes over the next 20 years, he should buy a Lycoming for his project. If he really wants to build his own PSRU for an auto engine, he needs to do that and forget about going flying whenever the urge takes him. Too many guys have spent vast sums on such things and in the end the had to stick an aircraft engine on the airplane just to stop the financial hemorrhaging. Auto conversions are not cheap; if they are, they're not reliable. Following the stories for 37 years teaches one something, as does actually installing and flying a Soob and PSRU on a Glastar. Give me the Lyc next time.

Dan

Speaking of tractors, some tractors have friction drive. Friction drive could be a small steel drive wheel running against a larger rubber driven wheel. This would be durable and not critical with regard to alignment. And some slippage might be good.

NASA did a study about friction drive for helicopters. I think it is called "tribolic drive" or something like that. Would this still have torsional problems?