Torsional Vibration and Resonance - Basic Theory and Issues

Discussion in 'General Auto Conversion Discussion' started by wsimpso1, Nov 20, 2012.

Help Support HomeBuiltAirplanes Forum by donating:

  1. Apr 27, 2019 #101

    plncraze

    plncraze

    plncraze

    Well-Known Member HBA Supporter

    Joined:
    May 12, 2006
    Messages:
    1,442
    Likes Received:
    284
    Every time I think of dropping my subscription they put in something like Anders or "Stressing Structures."
    It's like they read my mind LOL.
     
  2. Apr 27, 2019 #102

    Heliano

    Heliano

    Heliano

    Well-Known Member

    Joined:
    Dec 24, 2015
    Messages:
    49
    Likes Received:
    24
    Location:
    Campinas, SP, Brazil
    I am joining this thread a little late. I've read the postings since page one, though. Congrats, Billski. Very good arguments. And the whole discussion is high quality. This is how I view it:
    To design an engine reduction device attached to a propeller, good engineering and good testing is a must, be it belt, chain or gear. Any configuration can work out ok, provided enough analysis and testing is done. Some numbers: for example: a four-piston engine fires every 180 degrees of shaft rotation, and torque output varies roughly from 50% to 400% of the average torque along 180 degrees of shaft rotation! This torque oscilation itself is a big issue. Piston engines are called "intermitent combustion" engines, as opposed to "continuous combustion" such as turboprops. Is is a lot easier to develop a PSRU for a turboprop than it is for a piston engine for that reason. But turboprops have devices that can very well detect an impending PSRU failure well before it happens: they have chip detectors, oil temperature sensors and some of them have vibration sensors. I wonder why those developing PSRU's for piston engines are NOT considering such safety devices.
    Modeling the dynamics of a PSRU - and nowadays there are excellent tools for numeric (Finite Element) dynamic analysis such as ANSYS, NASTRAN, FEMAP, you name it. But the quality of the analysis is as good as the model you input. A good model, given the complexity of the dynamic scenario, is not easily built. To certify an engine, analysis is not enough. FAA (CFR14 Part 33) still requires MANY test stand hours with a lot of sensors producing data before the flight tests are done, no matter how good your finite element analysis is.
    However I've noticed that many (not all) existing PSRU's are not backed by sound engineering, especially belt units. This forum already discussed the problems related to belts and the need for an idle tensioner pulley, but numbers are needed.
    As for gear units, there is the never ending discussion about helical gears vs spur gears, but I think that this is NOT the most important issue. The most important issue is how to cope with the intermitent combustion. I see the solution adopted by Viking Aircraft (A BMW elastic cardan coupler) as a good one. It is NOT a torque damper. Rather it is a torque absorber, because it does not dissipate any energy by friction or any other means.

    Sorry for the long text.
    Heliano
     
  3. Apr 27, 2019 #103

    pictsidhe

    pictsidhe

    pictsidhe

    Well-Known Member

    Joined:
    Jul 15, 2014
    Messages:
    5,890
    Likes Received:
    1,500
    Location:
    North Carolina
    Heliano, something else to consider on many engines is the reciprocating weight. Singles, even fire twins and fours all accelerate their pistons simultaneously. That causes a significant torque variation. It is double rotation frequency. Many engines do cancel this out, ignoring the higher order effects from the finite conrod length. 90 v engines without offset pins , triples, fives, radials. Some of the offset pin engines may do too.
     
    wsimpso1 likes this.
  4. Apr 27, 2019 #104

    geosnooker2000

    geosnooker2000

    geosnooker2000

    Well-Known Member

    Joined:
    Mar 30, 2019
    Messages:
    71
    Likes Received:
    7
    Location:
    Somerville, TN
    As I was weed-eating this afternoon, I had a thought. I don't know, maybe this is common knowledge, maybe not, but practical experience just taught me something. As I tapped the line trimmer on the ground, I felt something I have felt a thousand times before, but never paid it any attention. Increased vibration. Now, true enough, my trimmer only has one line instead of two coming out of the head, but I have owned a gas trimmer back in the days of 5 acre property, and to a lesser degree, the same sensation was present. I'm about to make a leap of an assumption, so y'all just tell me I'm stupid if this is incorrect. Switching from a 2 blade prop to a 3 blade prop will reduce torsional vibration, NOT ONLY because of dividing the load by three instead of by two, BUT ALSO because of a reduced radius on the prop (just like my line trimmer was smoother, although ineffective, with short line).
     
  5. Apr 27, 2019 #105

    pictsidhe

    pictsidhe

    pictsidhe

    Well-Known Member

    Joined:
    Jul 15, 2014
    Messages:
    5,890
    Likes Received:
    1,500
    Location:
    North Carolina
    The answer is complicated.
    3 blade props usually have higher inertia than an equivalent hp 2 blade, though being a little smaller in diameter. There are a whole mess of factors influencing whether you have TV, one major one being the prop inertia. Increasing it may be good, but it may be bad...

    One advantage that 3 blade props do have over 2 blade props is in their gyroscopic behaviour. Lets say that you are spinning your plane, deliberately or not. the gyroscopic effect of each blade depends on the axes it is being turned around. Long story short, it's gyro effect varies with the sin of the angle relative to the spin axis. Two cycles per rev. For a single blade or twin blade prop, this results in the gyro effect of the whole prop being cyclic. With 3 or more blades, the effects of each individual blade cancel out when summed. The prop has a constant precessing torque.
     
  6. Apr 28, 2019 #106

    wsimpso1

    wsimpso1

    wsimpso1

    Well-Known Member

    Joined:
    Oct 19, 2003
    Messages:
    5,627
    Likes Received:
    2,866
    Location:
    Saline Michigan
    If you are speaking of opposed fours, yes. Inline fours generally have half of the pistons going up while half are going down at any one moment, balancing this vibe mode except in pitch axis of the engine - Engine guys call it rocking mode and is the reason a number of fours run balance shafts at twice engine speed.

    The biggest single torsional vibe input when running at serious torque is always firing order. The second biggest is usually at twice firing order, as the crank and con rod arrangement has different piston accels in the top and bottom halves of the rotation. A normally aspirated Otto cycle engine at full throttle will have firing order at one level, 2x firing order at about 1/3 to 1/4 the accel seen at firing order, and all the rest (1/2x firing, 4x firing, and others) are still smaller... I have run a lot of road vehicle powertrains and one Jabiru four cylinder airplane engine and they all have strong firing, next strongest 2x firing, and then all the rest are significantly smaller again. We tend to work most about 1x firing and only a little about 2x firing.

    In a flat four all four pistons are doing the same accel-decel dance together, and so while they have a 2x rev and 4x rev components, firing order is evenly spaced at 2x rev and it dominates deflections and resonance issues. We rarely get far into looking at the 2x firing order (4th rotation order) or others.

    In V8's and V10's, biggest order is firing order (4 or 5x rev), next biggest is 2x firing (8x or 10x rev), and then 12x/15x and 16x/20x. We could see some 2x rev in those engines, but it usually attributed nicely to cylinder to cylinder and bank to bank variations. Yeah, each cylinder and an engine with two banks has a clean (but small) 2x firing signature.

    While we know that the 2x rev component HAS to be there, when you take off firing and 2x firing orders, the rest of the spectrum is pretty low energy, and usually pretty unimportant in vibration isolation (soft systems).

    Billski
     
  7. Apr 28, 2019 #107

    TFF

    TFF

    TFF

    Well-Known Member

    Joined:
    Apr 28, 2010
    Messages:
    11,137
    Likes Received:
    3,017
    Location:
    Memphis, TN
    Lycoming with the front mounted prop governors have counter weighted crankshafts. Whatever order harmonic dampening. Those are the engines that three bladed props usually came on. The rear prop gov engines I think are all two blade in general. Lycoming has like a 1000 type Certificates pretty much all parts bin mix and match. I’m sure someone will find an example for me. Two blade has equal weight at each end. 4 cylinder airplane engine has crank that is equal moving in or out. Three blades have anywhere you cut the disk in half has one one one side and two on the other. Weights balance out, but they don’t phase the same.

    Two bladed helicopters can have a locked gear. Three or more bladed helicopters have to have a suspension like oleos. Eurocopter/Airbus have a special rotor head and semi compliant gear to solve the same problem but it’s solving a problem. If you don’t the helicopter will flip over in about one second as the blades come up to rpm. I have a friend with a Mooney. It had a prop strike. The insurance company wanted to put a three blade hartzel on, so he did. It was STCed. It always had a funny buzz no matter how it was balanced. He payed for the difference to put the two blade back on and it went away. It was a non counter balanced crank. I don’t really have an scientific explanation, but working with the stuff, it may matter for the combination. Three blade and six cylinder phases back again.
     
  8. Apr 28, 2019 #108

    pictsidhe

    pictsidhe

    pictsidhe

    Well-Known Member

    Joined:
    Jul 15, 2014
    Messages:
    5,890
    Likes Received:
    1,500
    Location:
    North Carolina
    In an inline 4, 2 of the pistons are being accelerated downwards while the other two are being accelerated upwards. As the throws of the up and down pistons are opposite, the crank thinks that they are all doing the same torsional thing to it.
    The 4 cyl cranks I've seen all had 1 throw up, 2 down, 1 up, giving perfect primary balance and no rocking couple.
     
  9. Apr 29, 2019 #109

    Heliano

    Heliano

    Heliano

    Well-Known Member

    Joined:
    Dec 24, 2015
    Messages:
    49
    Likes Received:
    24
    Location:
    Campinas, SP, Brazil
    The more I read in this forum the more I learn. Let me throw some more wood into the fire, as I do not have much knowledge about the inner workings of a piston engine, but know a litle more about aerodynamics: as pictsidhe mentioned, three-bladed propellers do have more inertia, but gyroscopic loads is another ball game. Another thing in favor of three-bladed propellers: aerodynamically-induced vibration (fundamental and harmonic frequencies) tend to be softer when compared to those induced by a two-bladed propeller (odd number of blades vs even number of blades). But again at the end of the day the thing that really matters is the numbers. When such a complex system (many inertia elements, aerodynamics and combustion-induced torque, coupler elasticity, all-interacting) is modeled and a numerical analysis is run (modal analysis and frequency response) more often than not the results are unexpected.
     
  10. May 24, 2019 #110

    Hot Wings

    Hot Wings

    Hot Wings

    Well-Known Member HBA Supporter

    Joined:
    Nov 14, 2009
    Messages:
    5,994
    Likes Received:
    1,999
    Location:
    Rocky Mountains
    Don't want to start another thread and this question is related.

    Thinking about a stiff system with about a 1 meter shaft length but would like to have a universal joint to take care of angular misalignment. Naturally the joint needs to be stiff as well. I see joints designed for this made with thin plates that are allowed to flex.
    Flexible disk couplings

    I also remember a Pinto that I swapped a motor in that ended up with about a .040' misalignment due to a starter shield overlapping the bell housing shield. (Manual engine into an automatic car) The torque plate only lasted about a week before it fatigued and broke at the crank flange.

    Standard drive shaft U-joint and CV joints don't seem like an option due to wear eventually leaving slop in the system.

    So.......
    Are there any universal joint designs, preferably off the shelf, that can handle a degree or 2 of misalignment and survive?
    Or, am I stuck with just having to use conventional joints and check the slop in the system at each inspection?
    Or, am I just over thinking this?
     
  11. May 24, 2019 #111

    Toobuilder

    Toobuilder

    Toobuilder

    Well-Known Member

    Joined:
    Jan 20, 2010
    Messages:
    4,287
    Likes Received:
    3,034
    Location:
    Mojave, Ca
    wsimpso1 likes this.
  12. May 24, 2019 #112

    pictsidhe

    pictsidhe

    pictsidhe

    Well-Known Member

    Joined:
    Jul 15, 2014
    Messages:
    5,890
    Likes Received:
    1,500
    Location:
    North Carolina
    A stiff system with that length shaft will need to be large. You should start by calculating how soft the drive can be, size the shaft for some fraction of that, then browse joint catalogues. A purpose designed flex plate could be the solution, though. Off the shelf joints tend to be designed for high torque in a compact size and will be very heavy once stiff enough.
     
    wsimpso1 likes this.
  13. May 24, 2019 #113

    Vigilant1

    Vigilant1

    Vigilant1

    Well-Known Member

    Joined:
    Jan 24, 2011
    Messages:
    3,555
    Likes Received:
    1,530
    Location:
    US
    You probably aren't overthinking it. But, if the fix to the slight misalignment is hard/expensive/risky to implement, it might be worth considering approaches to eliminate the misalignment by making the item(s) on one or both ends slightly repositionable (in location and angle). This eliminates the flex coupling and any extra supports it would require, and probably increases the torsional rigidity of the system.
     
  14. May 24, 2019 #114

    BBerson

    BBerson

    BBerson

    Well-Known Member HBA Supporter

    Joined:
    Dec 16, 2007
    Messages:
    11,419
    Likes Received:
    2,095
    Location:
    Port Townsend WA
    I am looking at "double roller chain couplers" at surpluscenter.com
    About $12 for each half. 73 to 194hp.
     
  15. May 24, 2019 #115

    radfordc

    radfordc

    radfordc

    Well-Known Member

    Joined:
    Feb 5, 2008
    Messages:
    1,151
    Likes Received:
    434
    Valley Engineering designed PSRUs for the VW engine and also for the Generac engine. At 5:30 in the video the designer describes an "overrunning clutch action" with a spring loaded "slack side idler". This design allows the engine to drive the prop, but the prop can't drive the engine. He states that this cancels one half of the resonance cycle preventing destructive resonance.

     
  16. May 25, 2019 #116

    Hot Wings

    Hot Wings

    Hot Wings

    Well-Known Member HBA Supporter

    Joined:
    Nov 14, 2009
    Messages:
    5,994
    Likes Received:
    1,999
    Location:
    Rocky Mountains
    This is kind of what I was thinking about. If it works in that aviation environment than it should work for me.

    Yes, but I have the space and the weight is looking to be less than a soft system. If I went soft I could just use the rubber driveshaft donuts like a lot of others have done.

    3 inch diameter steel shaft, with a WAG at some other details, and a middle of the road prop inertia gave about a 1500 rpm cushion.

    This is direct drive, not a PSRU. I'm thinking about mounting the whole engine, shaft and prop in a tube frame with that assembly mounted to the air-frame on rubber.
    Yes, I know that complicates things some more but from a 'build it easy' point of view it may be worth the trouble.
     
  17. May 25, 2019 #117

    wsimpso1

    wsimpso1

    wsimpso1

    Well-Known Member

    Joined:
    Oct 19, 2003
    Messages:
    5,627
    Likes Received:
    2,866
    Location:
    Saline Michigan
    Thread drift - geosnooker2000 is talking about a rotating element that is out of balance. The thread is about torsional vibration which rotating elements speeding up and slowing down rhythmically.

    Billski
     
  18. May 25, 2019 #118

    wsimpso1

    wsimpso1

    wsimpso1

    Well-Known Member

    Joined:
    Oct 19, 2003
    Messages:
    5,627
    Likes Received:
    2,866
    Location:
    Saline Michigan
    Thread drift.
     
  19. May 25, 2019 #119

    wsimpso1

    wsimpso1

    wsimpso1

    Well-Known Member

    Joined:
    Oct 19, 2003
    Messages:
    5,627
    Likes Received:
    2,866
    Location:
    Saline Michigan
    Flexible plate drives are used in automatic transmission systems between crank flange and torque converter. They take up axial, radial, and angular alignments due to tolerances all the parts in the system. These are deliberately made a soft to axial, radial, and angular errors, but are relatively stiff torsionally. They are specifically designed and developed around engine and torque converter dimensions and the loads of specific systems. You would have a development project on your hands.

    In order for it to be a stiff system, you need to raise the 1st torsional order to at least 150% of max firing frequency, with 200% being smarter. I have worked on a four cylinder project with a 44" shaft, and it was a soft system with the shaft as the softest element. You will need an order of magnitude stiffer shaft plus everything else must be similarly stiff.

    Other issues are that the engine must be able to do all of its motion on its mounts, the boom with the prop on it must be able to respond elastically to loads on it, and the prop will put in both thrust and gyroscopic moments while maintaining very large torsional stiffness. And as a "Stiff" system, you will not be able to tolerate much if any lash in the system, like splines. To be a stiff system, the prop must accelerate and decelerate with the crankshaft.

    This combination will be difficult to achieve, but might be possible with large universal joints at the engine and at the prop, plus an axially elastic coupling to allow for engine and airframe movements.

    I suspect that a much lighter system is possible with an elastic coupling between engine and shaft, and a deliberately soft shaft. Another possibility is to install a low spring rate isolator between the engine flywheel and PSRU to isolate firing order vibe from the rest of the system.

    Good luck either way.

    Billski
     
    Last edited: May 25, 2019
  20. May 25, 2019 #120

    wsimpso1

    wsimpso1

    wsimpso1

    Well-Known Member

    Joined:
    Oct 19, 2003
    Messages:
    5,627
    Likes Received:
    2,866
    Location:
    Saline Michigan
    Stiff, but with lash on torsional vibe reversals. Might be OK downstream of a soft element.
     

Share This Page

arrow_white