Chevy's

[DanH]

I like the Hessenaur article because it demonstrates how even very bright people can get their butts whipped in an effort to design a good torsional system by experiment only. On the flip side, it is kinda sad because all the information they needed was readily available.

Here is THE bible:

Ker Wilson, "Practical Solution of Torsional Vibration Problems", published by John Wiley and Sons, New York. It is a three volume set. Vols I and II, 2nd edition, date from 1940 while Vol III was published in 1965. "Practical Solution" is out of print, but you can borrow it from a large university library. Charles Fayette Taylor called it "the most complete treatment of the subject available in the English language".

Another good one is Nestorides, "A Handbook of Torsional Vibration", published in 1958. Again you'll have to borrow from a university library.

The best layman's quick reference for all things vibrational is J.P.DenHartog,
"Mechanical Vibrations", first published in 1934. You can buy a Dover reprint of the 1956 edition on Amazon for about $15. Cheap, and a great reference for your shelf. It is not specific to torsional vibration, although it does offer good coverage of the Holzer method.

Point is, all the fundamentals have been in place more than 70 years.

Dan

 

[rv6ejguy]

<<I'm not sure why a drive manufacturer like Eggenfellner or PowerSport would publish such data as these are only used on their specific engines and with the props they recommend.>>

Did a market survey of sorts a few years ago, as I was kicking around the idea of buying a telemetry rig. The question (to paraphrase) was "What would you be willing to pay for a torsional telemetry report on your engine/redrive combination?". The majority of respondents felt it was worth $1000. I think a $1000 premium would be a pretty good incentive if I were a redrive manufacturer.

The expensive firewall forwand package vendors might consider "conquest" customers, folks who currently stick with conventional engines because they don't accept offhand assurances. I'm satisfied that my new IO-390 will be happy paired with a 74" Hartzell BA because (1) the Lycoming engineers know precisely why they put pendulum absorbers on the crank, and (2) the Hartzell guys did telemetry. Neither shared the actual data with me, but I'm OK with that. I am SURE they have real stress values. I am not sure the conversion vendors have any measured values. Want my money? Show me.

Remember, I like auto conversions. I just think folks are accepting a woefully low level of real engineering, and it is holding back progress.

Dan Horton

Dan, sounds like you have a good business venture in the making as a TV consultant. The point is in your post, Hartzell and Textron are multi million (billion) dollar companies. Testing like this is a drop in the bucket for them. For someone like Eggenfellner who offers only a few, engine/ prop/ redrive combos, this may be money well spent if they haven't already done this testing but I'm pretty sure they did in conjunction with MT. Lyco produced many engines when combined with certain props had placarded rpm avoid zones. Many would say this was an engineering band aid and a cost cutting solution.

For the redrive only vendors, this testing is impossible as there are too many combos to test.

In the same vein, most of the popular drives have accumulated thousands or tens of thousands of hours of flight time. Flight time is better than all the theory in the world as it is real world with every real variable being applied in the actual working environment.

I absolutely agree that applying the proper theory in the design stage, checking with modern instrumentation and finally validating through extensive flight testing is the best approach. If you can get the people to spend that extra $1000 per drive, go for it.

As another business venture idea, build drives yourself and work on the design aspect in your marketing slant. If you can produce a better, tested drive than what is available now, believe me, you'll sell lots.

 

[Jess Meyers]

chevys

It seems the question is torsional vibration. I cannot speak for anyone but our units. It doesn't matter how you attach two independent shafts with gears, chains or belts there is some rpm in which they will "hook up" ie. a vibration point. There are so many equations to the problem, ratios, center distance's, prop weight , length, pitch.etc. Dr. Harrold did the analysis on our unit with various props, weights, lengths. We started with 72", 76" popular metal props, did the analysis at different pitch, and weights. The bearings we use in the front and the rear passed with loading to spare and extended hours of operation exceeding 1500 hrs. Included in this was the input from the crank pulse at ignition followed through to the exhaust cycle, This was for the V-6 of 262 cu. in. and the 350 cu. in. V-8. The V-8 with the pulse overlap is easier on the belt than the V-6. The V-6 calculations in hours of operations by far fell into the parameters of the belt life projections. We then proceeded to test the Warp Drive prop we are currently using, it too fell within the factors determined by the bearing and belt manufacturers. But the problem we have found is that if we produce a unit for 260 continuous HP some one says well I need 300 and so on. We had to call it somewhere so we said 260 continuous with the cooling factor that we have worked on for years. It all has a connection with other factors, ie. belt cooling as the lower sprocket will come to oil temperature over the operation of the unit, so cooling is an issue. Our units are designed to cool the upper portion of the engine, carb, distributor , with cool air with pressure in front of the radiator, and pressured air to the belt. One cannot just introduce air to the engine area without suffering from heating problems. As Dr. Harrold once told us you don't push a wagon with a rope, but pull it. So as with the radiator, pressurized air is in front of it but it requires a good exhaust at the cowl bottom to pull it through. Sorry about the length, but if this isn't the answer you want try to ask in another way and I'll try to answer it.
Jess

 

[gmcjetpilot]

What about testing

It seems the question is torsional vibration. I cannot speak for anyone but our units. Jess

So the answer is you did analytical calculations and flight test verify this. I guess DanH's post is that actual test sensors and equipment is available to test the torsional stain in the shafts. I think that is a good idea. I can't disagree. Any plan to flight test you drive with strain gages and get the real numbers. Also the good Doctor ran a bunch of equations. (I have a masters in engineering, mechanical so I am familiar with vibration analysis). Do you know if he used FEM, finite element analysis, or did he just do hand cranked calcs. Thanks in advance Jess.

George

PS: A few more Q's
How do you lube the bearings?
The bearing life is 1500 hours?
What is the cost to overhual the dirve?
How long to the belts last and what is their cost?

 

[Jess Meyers]

chevy

George, there were some of both, we built stands and took parts to destruction to verify the numbers Dr. Harrold obtained. The bearings are sealed double row ball rated radial and thrust. The belt life we recommend is 2 years or 500 hrs. whichever comes first, we have done actual testing exceeding this, our worry is someone leaving them uncovered outside exposing the belt to UV. The cost at present varies but it's under 200. it can be changed in less time than to change plugs on a Lyc. We try to take the very cautious approach. In the very early days we sheared a valve stem in the Swift and ended up in a ditch, it raised the belt adjustment over the largest amount and the belt was flown for 2 years more. Not good. But it still flew. The hit was so hard it bent the Al prop at the hub. The prop shop knew we had ruined an engine. The prop and part of the cowl were the only damage. The drive Overhaul cost including the belt is under 400.

 

[gmcjetpilot]

Thanks

George, there were some of both, we built stands and took parts to destruction to verify the numbers Dr. Harrold obtained. The bearings are sealed double row ball rated radial and thrust. The belt life we recommend is 2 years or 500 hrs. whichever comes first, we have done actual testing exceeding this, our worry is someone leaving them uncovered outside exposing the belt to UV. The cost at present varies but it's under 200. it can be changed in less time than to change plugs on a Lyc. We try to take the very cautious approach. In the very early days we sheared a valve stem in the Swift and ended up in a ditch, it raised the belt adjustment over the largest amount and the belt was flown for 2 years more. Not good. But it still flew. The hit was so hard it bent the Al prop at the hub. The prop shop knew we had ruined an engine. The prop and part of the cowl were the only damage. The drive Overhaul cost including the belt is under 400.

Thank you, very interesting, appreciate the stright answer.

 

[DanH]

<<It doesn't matter how you attach two independent shafts with gears, chains or belts there is some rpm in which they will "hook up" ie. a vibration point. There are so many equations to the problem, ratios, center distance's, prop weight , length, pitch.etc.>>

Thank you Mr. Myers.

Every drive has some torsional issues, and prediction only goes so far. So why not put a wheatstone bridge on the shaft, right there at the node behind the prop, and record actual vibratory and mean torque values?

I know the current "successful" redrives have lots of flight hours. True believers will argue endlessly about how good they are. The point is that it's well past time to make them better.

(1) The values will tell you the actual loads on the belt, the bearings, everything. No guesswork, no assumptions.

(2) With hard data you can lighten components. For example, check the photo of a propshaft earlier in this thread. It's a 2"D thinwall 4130 tube, and it would have sheared with the previous bad drive. No problem with the drive it serves now...check the telemetry graph. Any interest in a lighter system?

(3) Actual vibratory torque also tells you which RPM range(s) to avoid if you want to make it live even longer. Again check the telemetry example. Although it will handle a full throttle pass through the F1 peak (1500 RPM), in normal ops the pilot pushes up through it at part throttle and extends belt life. He also avoids that RPM range for continuous ops. Would your customers like such knowledge?

(4) You can make a design change, follow with a telemetry run, and know right then if you made things better or worse. There's no need to fly 1000 hours. Which is cheaper?

(5) Vibratory torque aside, the shaft telemetry is also a dyno, a bonus. Tie the tail to a truck, record mean torque at full throttle and bang your calculator. [(Torque/ratio) x RPM]/5252 = engine HP The little Suzuki in the above examples made 68.4 hp on it's best day and right at 64.0 on its worst. No guesswork. Would it be nice to see changes to your ignition or fuel systems?

So why doesn't eveyone do it?

Dan Horton

 

[Jess Meyers]

chevy

Dan we appreciate the input on TV for the units. We had in the past had to take a stand on the objective. 1. Power in the 150-180 range, make it reliable, make it cost effective. 25 years ago when we started this a lot of what your talking about was available only to HUGE Companies at great cost. Our original drive was for our own plane to test the power plant for our 2/3 scale ME-109E. At the start the drive was no problem it was all powerplant. We thought at one point we could finally answer the question why Auto engines won't work. But with persistance and Dr. Harrolds guidance we thrashed through the problems, it was just missunderstanding the set up. Now back to the TV. On 90 degree V-6 engines it is impossible to balance them throughout their entire operational range. We had to pick an area it was between 3-4000 rpm. We have found in the Island of Las Vegas that one engine can be balanced as many ways as there are vendors. And these are creditable shops. If we used the method and tried to go max. lightness on one engine it might fail while in another it would last forever. So the old factor if some is good more is better and too much is just right. If a company was building the entire product engine all the way through it could be made lighter but costlier.. We tried to make something very reliable, although over built, to fit in a cost range that the average person anywhere U.S.A. could find an engine and make him or her self a reliable 180 hp. and not have to refinance their house. Try not to pay more for your engine than your first house.
Jess

 

[rv6ejguy]

I know the current "successful" redrives have lots of flight hours. True believers will argue endlessly about how good they are. The point is that it's well past time to make them better.

(3) Actual vibratory torque also tells you which RPM range(s) to avoid if you want to make it live even longer. Again check the telemetry example. Although it will handle a full throttle pass through the F1 peak (1500 RPM), in normal ops the pilot pushes up through it at part throttle and extends belt life. He also avoids that RPM range for continuous ops. Would your customers like such knowledge?

(4) You can make a design change, follow with a telemetry run, and know right then if you made things better or worse. There's no need to fly 1000 hours. Which is cheaper?

Dan Horton

I think improvements could be made to every drive on the market- stronger, lighter, more reliable. Evolutionary changes have been made by every drive manufacturer I know of. Much of this comes from field experience and there is NO substitute for this.

I have two TV areas in my current setup- between 500 and 900 rpm and between 1200 and 1600 rpm. Idle is set at 1000-1100 and taxi range is 1700 and up. I can feel these periods easily. There are no periods in the flight range that I can feel although there may well be minor ones.

You certainly DO need to flight test drives. Any engineer who thinks that theoretical or ground testing will show up all problems is naive. All the multi billion dollar aerospace flops prove that. Nothing can duplicate the prop, air, thrust and gyroscopic loads, not to mention interaction with the engine under power, turbulence, G loadings and whirl mode dynamics.
Ever wonder why all certified aircraft undergo flight testing? This validates theory and design.

I wouldn't touch any redrive that hadn't accumulated at least 1000 hours of flight time, no matter how much FEA or TV studies were done on it.
Naturally the math and TV testing are a big plus, likely to save time and money in the future.

 

[DanH]

<<I have two TV areas in my current setup- between 500 and 900 rpm and between 1200 and 1600 rpm. Idle is set at 1000-1100 and taxi range is 1700 and up.>>

Would you mind a physical description of your engine/redrive/prop?

Dan

 

[rv6ejguy]

Engine in the RV6A is a Subaru EJ22 turbo flat four. Drive is a 1st Gen Marcotte M-300 (2.20 ratio) which uses an internal helical gear arrangement. Coupling is via the 1st Gen elastomeric bushings (6). I have modded the rear bearing as per Marcotte's instructions. The flywheel is the stock Marcotte aluminum 2 lb. piece. Prop is a 27 lb. composite 3 blade IVO Magnum.

The new RV10 project uses a Subaru EG33 twin turbo flat six and 2nd Gen Marcotte M-300 (2.04 ratio). Coupling is redesigned using higher durometer elastomers with reduced internal diameter in a revised drive plate design. The 2nd Gen also encorporates a ball bearing rear shaft support from the factory. Prop is a composite MT MTV-18-B/193-53A 3 blade C/S weighing 47 lbs.

 

[DanH]

Do you happen to know why Marcotte went to higher durometer drive bushings?

Dan

 

[rv6ejguy]

Yes, as mentioned previously, many Subaru installations experienced TV within the typical taxi and flight rpm ranges with the lower durometer bushings, especially with heavier props. There were some instances of the bushings becoming so hot that they failed. A friend of mine experienced serious TV in flight and removed the EJ25 and drive going to another brand with a Sprague clutch. The low durometer bushings actually excited the amplitude rather than damping it. Obviously in these cases, the energy has to go somewhere and the bushings get very hot.

In an effort to save weight, the flywheel is kept light which often causes problems as well. Eggenfellner changed to a heavier flywheel some years ago then to a dual mass/ damped flywheel similar to BMW.

High TV within the flight ranges is unacceptable. Many drive systems have ranges to avoid like Rotax 9 series engines around ground idle. This seems relatively common in my experience.

 

[DanH]

<<many Subaru installations experienced TV within the typical taxi and flight rpm ranges with the lower durometer bushings, especially with heavier props. There were some instances of the bushings becoming so hot that they failed.>>

How will harder bushings help?

Dan

 

[rv6ejguy]

The harder bushings appear to have moved the harmonic outside the typical 3000-5000 rpm flight ranges where problems were initially experienced. The soft bushings were "winding up" excessively upon torque impulses, interacting with gear backlash and the prop MI.

 

[DanH]

<<The harder bushings appear to have moved the harmonic outside the typical 3000-5000 rpm flight ranges where problems were initially experienced.>>

And this problem was experienced with the 4-cyl engine?

Dan

 

[rv6ejguy]

Yep. On EJ22 and EJ25 engines.

 

[DanH]

<<I have two TV areas in my current setup- between 500 and 900 rpm and between 1200 and 1600 rpm. Idle is set at 1000-1100 and taxi range is 1700 and up.>>

Ok, peaks at 700 and 1400 RPM, and you pass through 1400 RPM with each throttle-up and return to idle.

<<I can feel these periods easily.>>

As expected. The one that peaks at 1400 RPM seems to be the most powerful, yes?

<<There are no periods in the flight range that I can feel although there may well be minor ones.>>

Can't assume you'll feel any of the critical intersections above F1....that's the problem.

<<many Subaru installations experienced TV within the typical taxi and flight rpm ranges with the lower durometer bushings, especially with heavier props. There were some instances of the bushings becoming so hot that they failed.>>

....which would indicate operation at or near a critical.

<<The low durometer bushings actually excited the amplitude rather than damping it.>>

The bushings (like belts) are a connecting stiffness. They have no capacity to excite anything.

All excitement (power to force vibration) in this application stems from the various engine events that speed and slow crank rotation. Since they happen at regular periods we can assign frequencies to them and (no surprise) call them "exciting frequencies". The most significant is firing frequency, or as the geeks say, the "gas pressure oscillation frequency". Very simple math yields a list of exciting frequencies for a particular engine.
For example, firing frequency for a 4-stroke is just (RPM x #cyls)/120 = hertz.

The system has a number of natural frequencies, the actual number being inertias-minus-one. We label them F1, F2, F3, F(etc). We're not real interested in anything above F3 in this game. Determining natural frequencies takes far more math, and quite often requires measurement of stiffness and inertia values to guarantee meaningful input.

When a system natural frequency is matched by an exciting frequency, the oscillation gains amplitude. This is a "critical intersection" or a "critical speed". The intersection thing becomes clear when you plot both natural and exciting frequencies on a Campbell diagram.

<<Obviously in these cases, the energy has to go somewhere and the bushings get very hot.>>

The fact that the bushings got hot illustrates that they do have a damping function, but it also illustrates why rubber makes a very poor damper. "Damping" in correct context means to remove energy from the system, usually through heat. Rubber doesn't tolerate heat very well. We want to use rubber as a low-stiffness connecting element, not as a damper. Big difference.

<<The harder bushings appear to have moved the harmonic outside the typical 3000-5000 rpm flight ranges where problems were initially experienced.>>

Assume 1400 RPM to be the F1 at 46.6 hz with the soft bushings. An F2 in the 3000-5000 range is entirely possible. A harder bushing (a stiffer connecting stiffness) would raise natural frequencies, indeed a way to push that F2 above the cruise operating range. However, it would also push the F1 upward. A 4-cyl guy might have to taxi at less than the critical RPM rather than just above it as you do now. He would also face higher vibratory torque when he passes through it on the way to full throttle, since the intersection now falls higher up the engine torque curve (exciting frequency is more powerful). These are the choices, the art of intelligent compromise.

What about your new 6-cyl? Let's suppose the the harder bushings pushed the system F2 up around 200 hz. That would intersect with firing frequency at 6000 RPM in the 4 cyl application, well above the operating range. However, a 200 hz F2 would intersect with firing frequency at 4000 RPM in the 6-cyl application. Clearly you don't want that.

Believe it or not, the correct bushings for your new 6 cyl may be the softer ones.

With torque telemetry none of this would be guesswork <g>

Dan Horton

 

[rv6ejguy]

Actually the 700 rpm one is noticeably more prominent than the one at 1400 although both are places you don't want to be except transitionally.

My correction on the rubber dampers, as you point out they are more for coupling than damping but also provide some degree of damping.

I have to admit that for the first shot, I was lucky that there are no serious intersection points within the normal operating range. Fortunately this drive does give the choice of changing durometer of the couplers, with most drives these are few options other than changing the flywheel mass.

This has been interesting. I'll let you know how the six works.

 

[DanH]

<<Actually the 700 rpm one is noticeably more prominent than the one at 1400...>>

That's interesting. Unfortunately I can't model your drive without defined inertia and stiffness values. Gathering those is most of the work. With values in hand, Holzer-method software is available (cheap) to do the heavy lifting for F1, F2, etc.

<<although both are places you don't want to be except transitionally.>>

Right on brother.

<<the rubber dampers, as you point out they are more for coupling than damping>>

Folks often use the word "damper" to describe all kinds of couplers. Drives me crazy <g> A torsional coupler is just a torsion spring. A perfect torsional coupler takes in oscillating torque and outputs mean torque. Nothing is perfect of course.

<<Fortunately this drive does give the choice of changing durometer of the couplers>>

I'm not sure how homemade pin-in-bushing couplers got so popular for PSRU's when the pros (industrial drive and ship propulsion) consider the concept obsolete or a less than optimum choice. It's not just the slide-on installation ease; you can get that with quite a few of the engineered torsional coupler systems. It wouldn't take much to install an off-the-shelf engineered coupler with multiple stiffness choices and known performance parameters. Lord, Goetz, and Centa all have good lines. My favorite is the Centaflex line distributed by Lovejoy. Take a look:

http://www.lovejoy-inc.com/catalog/Torsional.pdf

Selecting a particular coupler requires that you know, minimum, the total inertia of the crank/flywheel assembly and the total inertia of the propeller/shaft/gear assembly. Collecting that info isn't as hard as it might sound. You just hang the parts from a bifillar pendulum.

Dan Horton

 

[rv6ejguy]

I'm not sure how homemade pin-in-bushing couplers got so popular for PSRU's when the pros (industrial drive and ship propulsion) consider the concept obsolete or a less than optimum choice. It's not just the slide-on installation ease; you can get that with quite a few of the engineered torsional coupler systems. It wouldn't take much to install an off-the-shelf engineered coupler with multiple stiffness choices and known performance parameters. Lord, Goetz, and Centa all have good lines. My favorite is the Centaflex line distributed by Lovejoy. Take a look:

http://www.lovejoy-inc.com/catalog/Torsional.pdf


Dan Horton

Yes, I've seen some fairly simple couplers used in industrial applications which might be adapted for aircraft redrives. I think the rubber pin couplers are dirt cheap, easy to install but probably not the best for this application with a wide rpm range. There is an excellent discussion on planetary drives here:

http://www.rotaryaviation.com/PSRU Zen Part 1.htm

http://www.rotaryaviation.com/PSRU Zen Part 2.html

 

[DanH]

<<I've seen some fairly simple couplers used in industrial applications which might be adapted for aircraft redrives.>>

You're only interested in the ones that list a dynamic torsional stiffness.

<<I think the rubber pin couplers are dirt cheap, easy to install but probably not the best for this application with a wide rpm range.>>

Right. Check the Lovejoy material for the concept of precompressed rubber working in both tension and compression.

Note rubber, not urethane. Urethane has a memory. The pin compresses the urethane during a load cycle and it does not spring back at a rate anywhere close to the 25 to 200 hz we need. Measure one of your urethane bushings, squeeze it in a vise, remove and measure it's diameter, then time how long it takes to return to it's original diameter. Poor material for a torsional coupling.

Dan

 

[rv6ejguy]

TV testing

Dan,

This discussion about torsional vibration testing has me curious enough to try to gather some info on my RV6A/ Sube setup. Can something useful be accomplished with a triaxial accelerometer and a Fluke portable scope. What frequency response range of sensor would be typically used and what sort of acceleration levels would you expect? Trying to put something inexpensive together to see if anything really scary is present in my setup.

Are there any integrated systems out there that you'd recommend which can data log with an overlayed rpm signal which are reasonably priced?

 

[Jess Meyers]

Chevy

It's interesting to see how chevy engines got to Subaruuuus and Rotarys. While the thread goes to dampening and what not the Chevy's just keep flying, Very Interesting....

 

[Rotary10-RV]

It's interesting to see how chevy engines got to Subaruuuus and Rotarys. While the thread goes to dampening and what not the Chevy's just keep flying, Very Interesting....

Quite true Jess, Thread Hijack tends to be a real problem. When in a thread like this I try to keep my personal engine choices to myself and address the item at hand I'll only chime in off topic if I see incorrect or misinformed information being put out there . I like the chevy option. I like the LS-1 engine option, although your set-up is V-6 to properly stay on-topic. The V-6 packages better than the V-8 leaving more room for cooling. Have any of your customers tried a High performance option? I know you were promoting a stock engine system.
Cheers,
Bill Jepson

 

[Jess Meyers]

chevys

Bill, point well taken, I see your a rodder and cyclest. As they say speed cost's money how fast do you want to go? When we first flew the sixA in 96 the red line was vague, with a 4 bbl we were up in the 210 plus range, and quite frankly our plane is a slug, built in 90 days looks like 30. We were afraid of what might happen so elected to go back to the 2bbl. carb, we installed the mixture control and left it at that. The 4.3L is rated at 262 at 5200 so in effect it's quite derated as we use it. We went with Van's philosophy keep it simple, keep it a day VFR plane. I hope this answers your question, and with the Vari-Prop it is quite peppy by 180 stds.

 

[DanH]

It's interesting to see how chevy engines got to Subaruuuus and Rotarys. While the thread goes to dampening and what not the Chevy's just keep flying, Very Interesting....

Oops, sorry, let's talk about Chevys. Jess, are you saying that the Chevys are somehow immune to torsional issues?

Dan

 

[rv6ejguy]

Jess, does the Chevy have any avoid zones in the rpm/ map ranges with your present prop? My Sube and the Rotax engines have bad periods around 1200 rpm. What is the long block weight of the V6 compared to the Buick V8?

 

[Jess Meyers]

chevy's

As your aware any 90 degree V-6 cannot be balanced throughout it's entire rpm range, we picked between 3-4000 rpm. balanced it there and also have the prop matched to it in that range. Any time you connect two shafts wheather by chain, gears, or belts there is some point they "hook up" with the belt it can be an annoying hum. With one unit there was an rpm of 3450 20 rpm above or below it was gone. With the Vari-Prop we did not find the range it was in. Could have been the wood blades, their profile width length any one of these. As for the dry wt. the Buick was 262 lbs the Chevy cast iron 292, thats without carb, exhaust starter etc. bare weight. The cast iron blocks are lighter today, and the accessories are much lighter. Such as the starter, manifolds etc.
Jess

 

[DanH]

<<any 90 degree V-6 cannot be balanced throughout it's entire rpm range, we picked between 3-4000 rpm.>>

The subject of interest is torsional vibration, not engine balance.

<<Any time you connect two shafts wheather by chain, gears, or belts there is some point they "hook up" with the belt it can be an annoying hum. With one unit there was an rpm of 3450 20 rpm above or below it was gone.>>

A standing wave in the belt would again be a different subject. Can you to focus on torsional issues? Your system has critical periods (intersections of natural frequencies and exciting frequencies) just like any other. The practical issue is the amplitude of the resulting resonant vibration. What have you observed, calculated, or measured?

Dan Horton

 

[Jess Meyers]

Vibrations

Dan, if this is your interest, contact GM as they may have the data your requesting, we have not done testing on that issue.
Jess

 

[DanH]

Jess,
GM certainly did torsional modeling of their crank and flywheel, and of their crank and flywheel when coupled to the manual transmission powertrain. They would have done at least these two models because the system is run clutch engaged or disengaged. Your application is an entirely different torsional system.

With all due respect (I do appreciate your pioneering efforts), your posts make it look like you lack any understanding of torsional vibration as it applies to a crankshaft/flywheel/lower shaft/upper shaft system. Can that be true? I find it hard to believe.

If true, the only way you got a successful PSRU system was to keep adding strength (usually meaning weight) until things didn't break. Ignoring an effort to reduce the underlying vibratory loads (which are usually higher than engine oscillating torque) means you built a system that is somewhat heavier than necessary. The method works, but weight has been one of the classic objections holding back acceptance of auto coversions.

Didn't mean to hijack your thread. Lighter and more reliable propeller drive systems are possible. It's gonna take engineering and modern tools. I entered the thread to ask your opinion regarding those tools. I ending up pushing Ross because I know he's smart and has the resources. I hope you won't knock his efforts should he choose to pursue it.

Dan Horton

 

[Rotary10-RV]

Jess,
GM certainly did torsional modeling of their crank and flywheel, and of their crank and flywheel when coupled to the manual transmission powertrain. They would have done at least these two models because the system is run clutch engaged or disengaged. Your application is an entirely different torsional system.

With all due respect (I do appreciate your pioneering efforts), your posts make it look like you lack any understanding of torsional vibration as it applies to a crankshaft/flywheel/lower shaft/upper shaft system. Can that be true? I find it hard to believe.

If true, the only way you got a successful PSRU system was to keep adding strength (usually meaning weight) until things didn't break. Ignoring an effort to reduce the underlying vibratory loads (which are usually higher than engine oscillating torque) means you built a system that is somewhat heavier than necessary. The method works, but weight has been one of the classic objections holding back acceptance of auto coversions.

Didn't mean to hijack your thread. Lighter and more reliable propeller drive systems are possible. It's gonna take engineering and modern tools. I entered the thread to ask your opinion regarding those tools. I ending up pushing Ross because I know he's smart and has the resources. I hope you won't knock his efforts should he choose to pursue it.

Dan Horton

Dan, I don't know about Jess, but I do consider your comments cogent. What may also be a good idea is to start a TV thread in the alternate engines forum. I'm a Mechanical engineer myself, but have rarely needed to do a true tortional analysis. The aircraft drive is a perfect test case. The link to the Lovejoy site is also appreciated, as I am often fumbling around looking for a damper/coupling for lower power stationary mechanisms.
Bill Jepson