This is just engineering of the simplest type. First, the rear end of automotive crankshafts are intended to transmit torque at rpm and tiny axial/radial loads. The axial loads are clutch throwout bearing forces and the radial loads are starter gear reaction forces and due to tolerance level errors in crankshaft to engine shaft alignment. Automatic tranny flex plates are designed the way they are to keep axial loads down. No where did the engine guys at Subaru or anywhere else design the crank for running a propellor. PSRU design must take torque at speed from the crank hub, and isolate or otherwise apply all of the other loads.The thing that concerns me about PSRUs is that you are transferring all of that pulling force from the prop from the crank shaft to the output shaft of the PSRU. Bearings will be real important. Basically the whole weight of the airplane is being pulled by this one output shaft, not to mention the bolts that hold the PSRU to the bell housing of the engine. So basically the whole PSRU (casing and bearings) is to bear ALL of the thrust force. IDK... when you think about it, it's not as bad as holding the airplane up in the air hanging by the prop. The thrust force is only equal to the drag it is overcoming.... except for when it is climbing.
Turbocharged intercooled Subaru EJ series engines are the light way to a solid 200 hp airplane engine, but kind of heavy for a 140 hp engine. Consult with Ross at SDS and sdsefi.com on this.Thank you for the thoughtful post. I would agree with the VW being the choice, but I'm not going to try and fly a 4 seat plane behind a100hp or less engine. I'm looking for 140hp, minimum. That is why I'm looking at the Subaru as an option. It is the right hp range by all accounts.
So by-pass a lot of this and go for a new set of problems just use the engine to power a generator and use the generator to power a electric motor. The result will likely be at least for the first generations slightly heavier than a PRSU but look how much safer it is in regards the loads imposed on the crankshaft? Also modern electronics are generally much better than those of our by-gone era's, that is they are reliable and "look ma" no gears. The ratio of engine/generator speed to the motor speed can be established by both the the generator and the motor windings and magnets and is virtually infinite within the practical range of aviation needs. The heavier portion, the engine/generator can be placed in any convenient and logical location on the aircraft and the motor/propeller portion is lighter than a conventional engine and gives more flexibility to were and how it is located.This is just engineering of the simplest type. First, the rear end of automotive crankshafts are intended to transmit torque at rpm and tiny axial/radial loads. The axial loads are clutch throwout bearing forces and the radial loads are starter gear reaction forces and due to tolerance level errors in crankshaft to engine shaft alignment. Automatic tranny flex plates are designed the way they are to keep axial loads down. No where did the engine guys at Subaru or anywhere else design the crank for running a propellor. PSRU design must take torque at speed from the crank hub, and isolate or otherwise apply all of the other loads.
PSRU design just starts at sizing shafts and bearings for gear loads and prop forces/moments. Propellor hubs have been standardized, prop shaft sizes/materials are well established. Beyond that you calculate the loads on all of this stuff and design shafts and housings, select bearings based upon loads and speeds and desired lifetime/reliability. When you get to thrust forces, they really are not very big. Gyroscopic precession puts way more stress on a prop shaft. I of the prop times rotation speed of the prop times max combined yaw and pitch rotation speed.
We usually hang the PSRU on the RFOB (rear face of block) for a simple set of reasons - that is the end of the crankshaft for power delivery, and that is the end of the engine block for torque and moment reactions. In cars, big loads and moments get fed back into the bell housing and then the block, and the RFOB is designed for that. Gear train reactions back into the case and then the RFOB are from engine torque times the gear ratio. The RFOB in car engines must do that too - in a front wheel drive car, the total ratio between engine and output shafts is on the order of 16-20, so RFOB will stand the 2-3 times we use in PSRU.
If this stuff makes your head spin, wait until you get into the rest of the topic.
The hard part in PSRU design is torsional resonance of the system. Here is a primer on the topic. https://www.homebuiltairplanes.com/forums/threads/torsional-vibration-and-resonance-basic-theory-and-issues.14215/Once you get into all of this, it can drive resizing of the entire system as well as sizing the isolating elements and driving adjustments in flywheel sizing.
Engineering is part science and part art, but the science part has to be done right. Sounds like maybe you should let other folks do the engineering and then trust that choice.
All of the four auto conversions that I referenced have excellent records. Those were... VW / Corvair / Viking Honda / Aeromomentum SuzukiThe advice to look at AeroMomentum is good advice. I have spent quite a bit of time with the owner, he knows his stuff, he has been developing a good reputation with his turn key engines, and has quite a few flying.
There is another guy selling turnkey engines who has an awful reputation, and I do not know how he can even stay in business... look around the web for good and bad reviews on engine sellers.
I used to think that way, until one day I realized that some of the people I thought were smarter that I... were not......Engineering is part science and part art, but the science part has to be done right. Sounds like maybe you should let other folks do the engineering and then trust that choice.
Funny you should mention that, I was just checking out Zenith last night.All of the four auto conversions that I referenced have excellent records. Those were... VW / Corvair / Viking Honda / Aeromomentum Suzuki
Depending on your aircraft, mission, etc any of these will serve.
Based on real world installations on the Zenith engine forum pages, including ease of installation and customer support, for "turn-key" engines both Viking and Aeromomentum have been very well received.
Aeromomentum comes out of AirBoat applications which match aircraft continuous high rpm use. Viking has over 500 engines sold and a long record of improving their product via fuel system design, electrical design, installation support videos, etc.
Dozens of good comments on record for both companies. These two companies have different philosophies about what makes a "good" engine. Both are fuel injected engines, ECU controlled. Either would be a good choice.
Check out the Zenith forum for each of these engines. Affordable. Reliable. Good support. Why reinvent the wheel with a Subaru and have all the fuss and bother to get things right?
If this is really your mission then you are in the heart of Lycoming territory. Look for a low time 320/360 off a wind damaged Skyhawk or Cherokee, add EFI/ignition and be done with it. Compared to an auto conversion you will be ahead financially, your reliability risk drops a bunch and you will be flying YEARS earlier.... But I still need roughly 140Hp or better, so I'm not completely sold if Corvair can satisfy my mission (that being, 4 place, cross country, 250lbs+/- luggage, 150Kts or better cruise).
The real problem is that you can not evaluate the whole design except by seeing if ALL of the customers have been happy, or how many are particularly unhappy. I am a retired transmission engineer, with considerable expertise on this topic, and I do not dare pronounce a product acceptable from their spec sheets and sale documents. I judge products I am buying based on how well they have been working in the field. For a powerplant, I want to know how many are flying and how the service history has been running. PSRU's are particularly tricky as they are vibration isolator, load isolator, gear box, and prop mount all rolled into one. One little detail not completely right, and you are in a glider, maybe on fire, maybe with critical parts busted by flying pieces of the engine, prop, or gearbox, maybe shaking the engine right out of its mountings... One little detail can sink the whole ship.I used to think that way, until one day I realized that some of the people I thought were smarter that I... were not.
I check most everything now.
Just to let you know, I have a masters in Architecture, which included 3 semesters of structural engineering, including wood, steel, and concrete (surprisingly, concrete was the most complicated). So I have a pretty good grasp of the engineering process. I am just moving into a new realm, and am feeling my way around to figure things out. I'm not saying I'm gonna design my own PSRU or 5th bearing. What I AM SAYING is, I want to learn enough to make an informed decision on my powerplant choice, the components I choose, the makers of said components, etc.
I did that. An EJ22 into a Glastar, using an RAF redrive. Not my airplane; someone else's. In the end he could have had a nice Lycoming for less money, and I didn't even charge for the work I did. It's a LOT OF WORK and it's not cheap. Nothing about aviation is cheap. I had to build a prototype mount from inexpensive steel tube, and then a tacked-together 4130 mount that I had welded by some TIG experts. My welding experience is mostly stick and MIG. That mount had 17 pieces of tubing in it in order to reach around the engine and grab it at its business end. Way more complicated than a Lyc's Dynafocal mount. The engine needed mounting points machined, welded and installed, and I MIGged those. Same with the mufflers, of which I built SEVEN iterations, trying to get effective silencing in a really tight spot. Cooling required the building of a plenum for the radiator. The fuel and ignition and electrical systems....oh, dear. A pile of work.My business for the last 24 years has been designing, manufacturing, selling and supporting aircraft EFI systems. We've sold nearly 900 of these for auto conversions in aircraft worldwide and another 1100 for aviation engines like Lycoming, Continental, Rotax and Jabiru.
I've helped many of my customers solve a multitude of issues with both types of engines. Having been down the auto conversion route twice myself with 4 and 6 cylinder Subarus and helping all these folks, I can say that you need a serious background in engines, welding, machining and fabrication to have a hope of pulling this off successfully. Even then, many have failed and gone back to a traditional aircraft engine after years of frustration and failures. This is much harder to do than it looks at first glance.
Most who've been successful, have been engineers, machinists and mechanics. If you're not one of those, I can almost 100% assure you that you won't be able to make this work reliably and safely in any reasonable time span.
Build a plane so you can work on it yourself if that's your goal but don't attempt to roll your own auto conversion. Just trying to save you time, money and grief, based on my experiences.
I have followed the Viking Honda almost since inception. "Lot of Folks" is just not the fact.I would be really careful about doing business with Jan Eggenfellner and Viking. There have been a lot of folks pissed with Eggenfellner and both of his companies...