More Thoughts on PSRUs

Discussion in 'General Auto Conversion Discussion' started by rv6ejguy, Jul 4, 2014.

Help Support HomeBuiltAirplanes Forum by donating:

  1. Jul 15, 2014 #81

    raven-rotor

    raven-rotor

    raven-rotor

    Active Member

    Joined:
    Mar 6, 2012
    Messages:
    33
    Likes Received:
    21
    Location:
    Taos, NM
    Nineteen years now of working with the 3 and 4 cylinder Suzuki® engines with a proven cog belt design. Thousands and thousands of flight hours on a few hundred redrives and no torsional failures to date with all kinds of different propellers, aircraft, drive ratios, etc. IMO the 3 cylinder is the most challenging application. Our success is due primarily to my own stubbornness and a LOT of test stand time under real world prop loading before we went to market.

    The great thing about the belt drive regarding torsional testing is that it is totally visible during the resonance phasing. Watching the sinusoidal movement of the belt with different drive ratios, propellers, flywheel weights, and dampening elements in the drive system allows direct feedback of resonance phasing that sometimes cannot even be "felt" by touch. A strobe light in phase with ignition pulses can also reveal more useful information.

    Its great to see this topic being addressed, but in the end its a matter of real world testing.

    Hope this helps.

    Jeron Smith
    Raven ReDrives Inc.
    303-440-6234
     
    cheapracer, clanon, delta and 3 others like this.
  2. Jul 15, 2014 #82

    akwrencher

    akwrencher

    akwrencher

    Well-Known Member HBA Supporter

    Joined:
    Oct 16, 2012
    Messages:
    1,082
    Likes Received:
    377
    Location:
    Gustavus, AK
    LOL. I have a collection of tension and Idler pulleys from changing timing belts :) And my best bud owns the scrap yard. I also have a machine lathe and mill, although just getting it set up and need to learn some skills first. I was mostly thinking of using dual tensioners for the test rig, in the hopes I could visually see which rpm's had problems with vibration, maybe using a strobe light. Not sure if it would work, but worth a shot. Anything around here is better to do yourself if possible, it's a long ways and a lot of shipping $ to any kind of shop that could do that sort of thing. This whole project idea is mainly only worth doing for me because I want to learn more in these areas, machine skills, testing skills, etc. Maybe some basic electronics, like arduino and using my scope for other than chasing down problems on cars. If I come up with something good it will eventually go on a plane. Other wise, no great loss, as it's all fun and learning :)

    Reg v belts would not be my first choice. So hard to get them to run in pairs, let alone more than two. Hope to find a micro v that is wide enough to do the job. We will see. Prob wont get to making any metal shavings until winter, but at least I can ask questions and source bearings, etc.

    BTW, thanks for the input. This is turning into a great thread.
     
  3. Jul 15, 2014 #83

    Jay Kempf

    Jay Kempf

    Jay Kempf

    Well-Known Member

    Joined:
    Apr 13, 2009
    Messages:
    3,635
    Likes Received:
    913
    Location:
    Warren, VT USA
    Forgot strobes in my earlier post. Very good cheap and easy timing/deflection visualization tool. Most people are familiar with timing lights. Paint marks, targets, timing lights can help to visualize lots of dynamic modes. Doesn't get you out of doing the math and analysis.

    One thing that people don't realize about test stand running is that you can look at and measure airflow around the cowl and all the inlet and outlets in the static condition and to a certain extent characterize in flight dynamics. Simple tufting can show all kinds of things and allow some initial optimization. If only to investigate taxi cooling performance it is still worth it. If one is designing an aircraft from scratch and wondering about the firewall forward or aft design it would seem for a powerplane that that is the FIRST thing you should source, build, test and optimize before moving on to the airframe details. If your engine choice doesn't work your airframe may be irrelevant.
     
    rv6ejguy and akwrencher like this.
  4. Jul 15, 2014 #84

    stol

    stol

    stol

    Well-Known Member

    Joined:
    Feb 15, 2008
    Messages:
    497
    Likes Received:
    145
    Location:
    Jackson Hole Wyoming
    As others have said.... A V-8 is alot easier on TV / Harmonics and other motor induced forces...... I personally believe in the outstanding results from using The Original Fluidampr on the nose end of any crank, in any application... The viscous fluids do a great job at absorbing vibes, spikes and internal harmonics on any motor..... Also, the unit is heavier then any other balancer, but it is on the part closest to the firewall so additional weight is easiely managed... My flywheel is aluminum and IIRC it is around 4lbs.... I wanted it,as a flex plate is too fragile and the ring gear needs good support and most, if not all flex plates not attached to a torque converter will eventually fail..... IMHO..

    Ps... The Raven redrive guy is probably seen horrorendous TV and is able to work around the issues..... Nothing beats total time on a one off creation...
     
    delta, MARCVINI and rv6ejguy like this.
  5. Jul 16, 2014 #85

    MARCVINI

    MARCVINI

    MARCVINI

    Well-Known Member

    Joined:
    Apr 16, 2014
    Messages:
    295
    Likes Received:
    12
    Location:
    Pitanga/Paraná/Brazil
    WELL, "NUFF" SAID!!! MY **** FOR YOU!!! (See that I´ve found my own "quad" trick? :gig:)

    Cheers.
     
  6. Jul 16, 2014 #86

    stol

    stol

    stol

    Well-Known Member

    Joined:
    Feb 15, 2008
    Messages:
    497
    Likes Received:
    145
    Location:
    Jackson Hole Wyoming
    I am not sure what a "quad " trick is... But . My only answer is........... "XXXX".....:roll:
     
    rv6ejguy and MARCVINI like this.
  7. Jul 16, 2014 #87

    MARCVINI

    MARCVINI

    MARCVINI

    Well-Known Member

    Joined:
    Apr 16, 2014
    Messages:
    295
    Likes Received:
    12
    Location:
    Pitanga/Paraná/Brazil
    Well, it is just a sign of respect and appreciation...
     
  8. Jul 16, 2014 #88

    MARCVINI

    MARCVINI

    MARCVINI

    Well-Known Member

    Joined:
    Apr 16, 2014
    Messages:
    295
    Likes Received:
    12
    Location:
    Pitanga/Paraná/Brazil
    Some time ago, driven by the idea that Hy-Vo chains behave the same way cog belts do, I contacted the owner of N427VA, wich is a quite fine stroker SBC V8 powered Velocity, and asked him whether or not a full weighed flywheel was necessary in his Hy-Vo chain PSRU. His answer was that he had already tried a lighter flywheel but the chain didn´t stand the engine´s pulses. It seems that his PSRU follows the same recipe as Geschwender´s.

    That is a very succesfull auto conversion, to say the least, is worth looking at and seems to confirm that, regardless of engine type, "metalic" reductions seem to be more dependent on flywheel weight than cog belts do. And that is something worth considering, since most auto conversions tend to be more on the heavy side.

    Here is the before mentioned "Beast":

    N427va_engine.jpg

    Marcvini
     
    Last edited: Jul 16, 2014
  9. Jul 16, 2014 #89

    rv6ejguy

    rv6ejguy

    rv6ejguy

    Well-Known Member

    Joined:
    Jun 26, 2012
    Messages:
    3,749
    Likes Received:
    2,785
    Location:
    Calgary, Alberta, Canada
    I disagree. The Rotax 912 has no flywheel to speak of on the output end of the crankshaft, nor do P&W, Wright, RR, DB, BMW or Allison engines... What's that, something over 500,000 engines in total...
     
  10. Jul 16, 2014 #90

    raven-rotor

    raven-rotor

    raven-rotor

    Active Member

    Joined:
    Mar 6, 2012
    Messages:
    33
    Likes Received:
    21
    Location:
    Taos, NM
    Marcvini

    'Metallic' reductions should not be painted with the same brush. There are fundamental differences in a gear set vs. a chain. Once you have studied the movement of a belt drive like ours, you have an appreciation for the moments that can me generated in a chain drive with its much larger mass. A belt drive simply cannot generate those slinging forces. Look at the sophisticated plastic guiding systems now used to contain this on all the late model chain camshaft drives and they are so much lighter than a HiVo. Gearsets do not have that issue.

    Jeron Smith
    Raven ReDrives Inc
    303-440-6234
     
    rv6ejguy likes this.
  11. Jul 16, 2014 #91

    MARCVINI

    MARCVINI

    MARCVINI

    Well-Known Member

    Joined:
    Apr 16, 2014
    Messages:
    295
    Likes Received:
    12
    Location:
    Pitanga/Paraná/Brazil
    Your input is correct. BUT the examples you mention, correct me if I am wrong, are aeronautical engines, with internals like bore & stroke thought out to favor reduction couplings.

    The Rotax is the most simbolic one, since it has very tiny stroke and, therefore, both its mean piston speeds and stroke pulses are "light". I believe that the shortest the stroke, the lowest are the torque pulse peaks, so, this engine, in particular, puts out power very "evenly" on its PSRU.

    That said, I believe I would better explain my point if I said that my statement aplies to the "average" auto engine. And your own instalation seems to confirm that, since you say that it seems to like the heavier flywheel you have instaled.

    The comparison between Stol´s belted set up and N427VA´s hy-vo set up, both being V8 engines, seems to confirm my statement, since Stol uses a 4 lbs flywheel and N427VA has to use a full weigh flywheel, wich weighs in more than 20 lbs.


    Marcvini
     
    Last edited: Jul 16, 2014
  12. Jul 16, 2014 #92

    timberwolf8199

    timberwolf8199

    timberwolf8199

    Well-Known Member

    Joined:
    Sep 3, 2011
    Messages:
    276
    Likes Received:
    64
    Location:
    Grand Rapids area, MI, USA
    FYI - To calculate drive capacity of a flat friction driven belt (not cogged or V) you can use the capstan equation: Th/Tl=e^(uB) where Th=high tension side, Tl= low tension side, u=coefficient of friction, B=effective angle of wrap in radians. Effective angle is ~10-15deg less than geometric angle. Using this will let you calculate the tension differential you need to transmit your power and from that you can derive a pretension value. With those forces calculated you can begin to size supports.

    A positive drive (toothed) or V-belt should be able to exceed these values if designed correctly. That said, calculating the capacity goes beyond shear strength and tensile strength...there's more going on there than most people (even some manufacturers) realize.
     
    cheapracer likes this.
  13. Jul 16, 2014 #93

    Jay Kempf

    Jay Kempf

    Jay Kempf

    Well-Known Member

    Joined:
    Apr 13, 2009
    Messages:
    3,635
    Likes Received:
    913
    Location:
    Warren, VT USA
    One great reason for cog belt cam drives is throttle response. The belt is a lighter thing to accelerate and decelerate. With the longer distances of multi valve per chamber overhead cam drives that becomes important. The reason most auto manufacturers are going away from belts is user generated abuse leading to warranty budget problems, not because of performance.

    Jeron, how much time did you spend messing around with idler setups before you arrived at your final redrive design?
     
  14. Jul 16, 2014 #94

    rv6ejguy

    rv6ejguy

    rv6ejguy

    Well-Known Member

    Joined:
    Jun 26, 2012
    Messages:
    3,749
    Likes Received:
    2,785
    Location:
    Calgary, Alberta, Canada
    Ahhh no. The examples I listed have a huge variation in strokes and bore to stroke ratios. Also, this has little effect on TV per se anyway and it's a myth that short stroke engines have lower torque as well.

    I am thinking you just don't understand engines or TV very well judging from your posts. There is a large variation in bore to stroke ratios in auto engines. For instance, Subarus have a large bore and short stroke, most Hondas are the opposite.

    ALL engines will have lower TV amplitudes using a flywheel with higher MMOI. As I explained before, after modeling my system, the addition of flywheel mass was the easier and cheapest solution to solving high amplitude TV at F1 and F2 on my installation. I have also explained that my friend has an identical engine and prop with an RAF belt drive and his F1 and F2 areas were at almost identical rpms. He will be adding some flywheel mass this summer also to help improve those areas.

    The variations in PSRU design, engine and damper type along with propeller MMOI are major determining factors in TV dynamics. Comparing 2 separate installations and concluding one needs a heavier flywheel because one happens to be a chain drive and one is a belt drive is essentially invalid IMO.
     
    wsimpso1 likes this.
  15. Jul 16, 2014 #95

    cheapracer

    cheapracer

    cheapracer

    Well-Known Member

    Joined:
    Sep 8, 2013
    Messages:
    5,110
    Likes Received:
    3,385
    Location:
    Australian
    Ross, putting the weight aside, any thoughts on the dual mass auto flywheels in regards to all this?
     
  16. Jul 16, 2014 #96

    MARCVINI

    MARCVINI

    MARCVINI

    Well-Known Member

    Joined:
    Apr 16, 2014
    Messages:
    295
    Likes Received:
    12
    Location:
    Pitanga/Paraná/Brazil
    Well. I have used Rotax 912 bore and stroke ratio just as an isolated and "illustrative" example among the many other "solutions" used in purpose built aeronautical engines, not implying that the "solutions" used on Rotaxes aplied to all other aeronatical engines you mentioned. I also didn´t say that this bore and stroke solution was the only one Rotax used in its engines to adress the issue under coment. Nor did I say that short stroke engines necessarily have lower torque: i only said that the "peaks" in the torque pulses are lower that those on longer stroke engines, that being one of the reasons the 912 goes so well with its psru. More strokes with less intensity are less likely to generate TV issues than less strokes with more intensity. That seems to be the case with 912s.

    I also haven´t said that belt drive PSRUs totally preclude the use of a flywheel. On most engines it will surely have to be there, specially ones with less cylinders. What I said is that real world experience seems to confirm that belt drive set ups are less dependent on flyweel weight to work right.

    Again, I believe that the comparison bettween the two V8 powerplants I mentioned I pretty valid, since they simply represent what happens in the real world, not only in theory, and both engines are similar in terms of TV behavior. Mr. Geschwender advocated the necessity of using full weight flywheel in his set ups, wich also seems to confirm my point.

    As for the majority of succesfull "reduced" aero engines being geared ones, it is valid if one consider purpose built aeronautical ones, the same not being true for auto conversions.

    Finaly, as for my understanding or not engines or TVs, well, I believe that this forum and thread is an open space for debates and I am totally open to knowing new things. And the way for me to learn new things, and I am not afraid to do it, is by expressing my opinions, even if they prove to be wrong. I believe that Jim Price (Long EZ jockey and aspirated engines world altitute record holder) is just right when he says that "the only stupid question there is is the one that is not asked". IMHO



    Marcvini
     
    Last edited: Jul 16, 2014
  17. Jul 16, 2014 #97

    MARCVINI

    MARCVINI

    MARCVINI

    Well-Known Member

    Joined:
    Apr 16, 2014
    Messages:
    295
    Likes Received:
    12
    Location:
    Pitanga/Paraná/Brazil
    The Belted Airpower guys say that the fact that their reduction ratio is close to an irrational number has a positive efect on their instalation´s harmonics. Anyone has any comment on that, or can confirm that?

    Marcvini
     
  18. Jul 16, 2014 #98

    rv6ejguy

    rv6ejguy

    rv6ejguy

    Well-Known Member

    Joined:
    Jun 26, 2012
    Messages:
    3,749
    Likes Received:
    2,785
    Location:
    Calgary, Alberta, Canada
    Your posts are beating on the premise that all belt drives can get away with a light flywheel where gear or chain drives cannot- completely false assumption with the examples I've given.

    A short stroke engine will have very similar torque impulses to a long stroke one of the same displacement. Connecting rod length to stroke ratio has more effect on this. Aircraft engine bore to stroke ratios are similar to lots of automotive engines.

    The 912 still has a TV issue like most 4 cylinder engines in the 600-1200 rpm range despite the short stroke and spring and ramp damper.

    How do you know anything about the TV signatures of the two V8 setups you mention since neither had TV studies done?

    Opinions can be based on fact or not. Debate without some facts is a waste of time IMO though. A belief that something is true does not make it fact- especially with no experience in the field.

    All drive types have pros and cons as I've said several times in this thread. If you prefer belt drives, that's great. Go get the parts and build it, stick it on your engine and start testing. When it works well for several hundred hours, we'll all applaud.

    If I was building my own drive in my shop, it would be a belt drive too but I wouldn't say it's better in every respect than a chain or gear drive. We've seen all 3 times used very successfully.

    Deny the physics and math if you want but the realities won't change. TV can still bite you with a LAR belt drive PSRU as dozens have discovered the hard way.
     
    Last edited: Jul 16, 2014
  19. Jul 16, 2014 #99

    MARCVINI

    MARCVINI

    MARCVINI

    Well-Known Member

    Joined:
    Apr 16, 2014
    Messages:
    295
    Likes Received:
    12
    Location:
    Pitanga/Paraná/Brazil
    We seem to be disconected but, in the end, all this discussion only adds to "the cause". I honestly appreciate all your addressings regarding my posts.

    Just one last thing (I promise I will not bother you again!!!): I didn´t say belt drives are better than chain or gear ones. What I did was only to "sugest" some advantages of the formers before the latters and doing it without disregarding the latters´ inherent advantages, whatever they might be. This is the reason why I tried to use the expression "seem to" as much as possible in my posts.

    That said, again, I thank you for your adressings.

    Marcvini
     
  20. Jul 16, 2014 #100

    Autodidact

    Autodidact

    Autodidact

    Well-Known Member

    Joined:
    Oct 21, 2009
    Messages:
    4,513
    Likes Received:
    799
    Location:
    Oklahoma
    Bother us, bother us! It's just a forum, and you're not a troll, so...:beer: better to have some conversation than none, IMO.

    The 912 crankshaft does have very large counterweights - relatively bigger than what you see on a VW crank, it looks like - and while that doesn't replace a flywheel, it does add some inertia; the point is that the elastic coupling needs be tailored to the inertias of the system, whatever they are.

    As for cog belts, I suspect that they may provide some damping, which is helpful; there is a lot of contact area on the teeth and that's an opportunity to generate some damping with a small amount of rubbing per unit area, which would keep wear down somewhat. Just a "pet' theory, though, and I'd like to see some more real data (in plain English) on these belts.
     
    MARCVINI likes this.

Share This Page

arrow_white