Belted PSRU question/idea

HomeBuiltAirplanes.com

Help Support HomeBuiltAirplanes.com:

akwrencher

Well-Known Member
HBA Supporter
Joined
Oct 16, 2012
Messages
1,167
Location
Gustavus, AK
Ok, I'm sure someone has probably already tried this, but I have to ask. Since one of the biggest chalanges with redrives is isolating them from the engine power pulses so they don't dissintigrate, why couldn't you use a cogbelt psru but with an idler pulley on each side pushing in using a spring. Wouldn't this allow the engine power pulses to be absorbed? Also, it would eliminate belt tension issues, and possibly reduce the overall side load on the crank bearing. Any thoughs/comments from the engineers out there? This is the idea I have wanted to try when I get to that point, so any info would be great.
 

Jay Kempf

Curmudgeon in Training (CIT)
Lifetime Supporter
Joined
Apr 13, 2009
Messages
3,865
Location
Warren, VT USA
It can be done. Tuning it would be an interesting investigation. I have been working on a design where there is an idler on both the drive and slack side of the belt. To do that they have to work together and be tuned both together and independently. That can all be done. I do not think that it will increase belt life but it would offer better control such that you don't lose life due to the system design. GTD belts actually have long life if used properly. To have an idler on the drive side it would need to be able to quickly be consumed so that the belt wouldn't be driving into the damper in the system. That would create problems instead of solving them. That drive side idler would merely be there to absorb slack on an overrun condition where the prop bounced ahead of the engine. With independent dampers you could tune all of those driveline frequencies. Most cog belt drives have a 180 turn slack idler now such that you can apply a spring pressure directly to the tension force in the belt. In general controlling the belt tension by just moving one or the other end of the assembly is not a good idea as it doesn't deal with thermal affects. As far as reducing overall belt tension on the crank, that would be true in a bad design but you still need to have the recommended tension to keep the cog belt anchored in the sprockets or you will jump or shear teeth. But it would do a better job of keeping tension even across all operating modes certainly.

The next step is to just build one which is where I am trying to get to. Not there yet myself. I have built a bunch of highly loaded and high RPM GTD belt rigs for testing of heavy rotating equipment. The engineering of these belts is pretty well known and the dynamics of prop and engine vibrations are also fairly well known. But the analysis is not even slightly trivial. Miss one variable and you have spontaneous disassembly.
 

DangerZone

Well-Known Member
HBA Supporter
Joined
Sep 5, 2011
Messages
2,110
Location
Zagreb HR
It can be done. Tuning it would be an interesting investigation. I have been working on a design where there is an idler on both the drive and slack side of the belt. To do that they have to work together and be tuned both together and independently. That can all be done. I do not think that it will increase belt life but it would offer better control such that you don't lose life due to the system design. GTD belts actually have long life if used properly. To have an idler on the drive side it would need to be able to quickly be consumed so that the belt wouldn't be driving into the damper in the system. That would create problems instead of solving them. That drive side idler would merely be there to absorb slack on an overrun condition where the prop bounced ahead of the engine. With independent dampers you could tune all of those driveline frequencies. Most cog belt drives have a 180 turn slack idler now such that you can apply a spring pressure directly to the tension force in the belt. In general controlling the belt tension by just moving one or the other end of the assembly is not a good idea as it doesn't deal with thermal affects. As far as reducing overall belt tension on the crank, that would be true in a bad design but you still need to have the recommended tension to keep the cog belt anchored in the sprockets or you will jump or shear teeth. But it would do a better job of keeping tension even across all operating modes certainly.

The next step is to just build one which is where I am trying to get to. Not there yet myself. I have built a bunch of highly loaded and high RPM GTD belt rigs for testing of heavy rotating equipment. The engineering of these belts is pretty well known and the dynamics of prop and engine vibrations are also fairly well known. But the analysis is not even slightly trivial. Miss one variable and you have spontaneous disassembly.
Just a thought... Wouldn't a progressive spring pulley on the slack side of the belt drive and a rubber cushion damper in the propeller sprocket be a solution with less moving parts, more reliability and more dampening?
 

Zoomzoooie

Well-Known Member
Joined
Feb 14, 2011
Messages
170
Location
BC Canada
I'm going to be making a cog belt PSRU and won't be using idler pulleys. I have seen some with and some without, mostly without. I think it depends on belt length between the pulleys. My pulleys will have about 5/8" space between each other.
My engine is a Teledyne a4084 - 45hp. These engines are also used on mini airboats with cogbelt PRSUs.

I don't have access to machining equipment and one of the problems I had was trying to find a low cost stub axle and came across the VW Jetta 80' - 98' rear stub axle which looks like it will work. From what little info I can find, they appear to be about 5" - 5.5" in length.
Before I get one I wanted to know if higher angle thrust bearing would fit, but I need the actual dimensions of the axle, which seems to be impossible to get. I think I just have to go ahead and get one and start from there.


I'll be adding a short shaft to the rear of stub axle where it and the mounting bolts will ride in slotted holes. There will be a tensioning bolt that goes throuth the short shaft to set belt tension. Once tension is set, the mounting bolts are torqued.
The pulleys will be 3.5" wide blower pulleys and an 85mm wide belt for best wear and pulse absorbtion. I need a 1.5 ratio for my 68" Ivoprop, so I will be using 8mm pitch 60T and 40T pulleys. I'm trying to design the PRSU so the pulleys will need little or no modification.

One thing I want to have done to the stub axle is to cross drill it and drill the end to the cross drill hole and put a grease fitting for easier grease maintance.
Anyone see any problems with having the axle drilled.

ZZ
 

akwrencher

Well-Known Member
HBA Supporter
Joined
Oct 16, 2012
Messages
1,167
Location
Gustavus, AK
Jay,

I assumed it would have to be tuned, spring tensions set to give desired characteristics. No idea how to go about that except trial and error which could be expensive. just today I was looking at a small Hughes two seat helicopter. It runs an io360 and the main clutch is eight 3/8 size v belts with an idler setup that puts tension on the belts when you flip a switch. Seams to me if it would work in a chopper that is certified, it could work on a plane? Has to be balanced of course, but that goes for anything you spin fast:) I kind of like the idea of using multiple v belts for lower (100 hp or less) power setups. In my humble experience, they dont take alot of tension from an idler to keep them from slipping, and also eliminates belt slap, plus with multi belt configuration, might it have less of a chance of sudden total failure?
Just some thoughts I have tumbling around in my head.......


ZZ, sounds like it sould work. I had a genset once with a cogbelt. Make sure you have a way to fine tune the driven pulley so that the belt will not want to "walk" sideways. Might have to do this after it is assembled and you can run it, as it is pretty precise. Shims could work, but would take a lot of trial and error. Just some food for thought, and good luck! :)

Zach.
 
Last edited:

Dan Thomas

Well-Known Member
Joined
Sep 17, 2008
Messages
5,230
An idler has weight, and therefore has a resonant frequency that might cause trouble when it starts bouncing against the belt out of phase with the firing pulses. Most likely the belt would start jumping teeth.

The PSRUs I've seen (and flown one) use no idlers at all. The prop shaft is on an eccentric that is turned to achieve the right tension, and locked there. Other might use other tensioning arrangements, but few will have idlers.

The helicopter V-belt thing is common and is found not only in the Hughes but the Robinsons, too. There's a difference, though, between these setups and a fixed-wing PSRU: the engine has a flywheel, and the transmission doesn't represent the same sort of inertia that a propeller does.

Still, v-belt PSRUs have been used for a long time and work OK. They lose some hp to efficiency losses, since the belt wedges in the sheave grooves and has to be pulled free. The bit of slip a v-belt might have will often save the system from resonant failures that plague other PSRUs.

Dan
 

deskpilot

Well-Known Member
Joined
Jun 30, 2009
Messages
1,121
Location
Morphett Vale, South Australia. Just south of Adel
Interesting subject, especially as I am considering dampers on my conta-rotation drive system. Contra-rotating props drive system.jpg This early model shows a belt for the primary drive but I went away from that idea due to my fears about bending belts backwards. Even so, a chain drive still needs to isolate engine pulses..or does it? Will my 3 gears take out the shocks? If not, then 2 idlers of different rating is my solution. A strong one on the drive side, and a weaker one to take up any slack due to chain stretch (from heat) etc. Yes, I am aware of the extra weight. I realise that my initial drive gear should be smaller for more torque.
 

akwrencher

Well-Known Member
HBA Supporter
Joined
Oct 16, 2012
Messages
1,167
Location
Gustavus, AK
Dan,
So the transmission effectively isolates the main and tail rotors so no resonant frequency can develop? I knew this was comparing apples and bannanas, but was worth throwing out there. i still think it might be worth pursueing, just for fun. Also, the idler does not have to be very heavy, and I would think that the spring tension would keep it from bouncing enough to let any real slack in the belts. I would think it would be best to use just one idler in a v belt drive, grooved pulley pulling out on the belt, on the lee side of the load, so to speek. This might work well with a single wide micro v belt also, as they tend to be real sensitive to belt tension, IME.

How much time do you have with a belt drive? Was it yours, or a freinds that you flew? I've heard so many different oppinions about belt drive reliability, I am hoping to get a little clearer picture from folks who have used them.

Thanks for your input, This is a great Forum!
Zach
 

akwrencher

Well-Known Member
HBA Supporter
Joined
Oct 16, 2012
Messages
1,167
Location
Gustavus, AK
Deskpilot, are you building a plane or helicopter? Sounds interesting. More info please:)
 

DangerZone

Well-Known Member
HBA Supporter
Joined
Sep 5, 2011
Messages
2,110
Location
Zagreb HR
I'm going to be making a cog belt PSRU and won't be using idler pulleys. I have seen some with and some without, mostly without. I think it depends on belt length between the pulleys. My pulleys will have about 5/8" space between each other.
My engine is a Teledyne a4084 - 45hp. These engines are also used on mini airboats with cogbelt PRSUs.

I don't have access to machining equipment and one of the problems I had was trying to find a low cost stub axle and came across the VW Jetta 80' - 98' rear stub axle which looks like it will work. From what little info I can find, they appear to be about 5" - 5.5" in length.
Before I get one I wanted to know if higher angle thrust bearing would fit, but I need the actual dimensions of the axle, which seems to be impossible to get. I think I just have to go ahead and get one and start from there.


I'll be adding a short shaft to the rear of stub axle where it and the mounting bolts will ride in slotted holes. There will be a tensioning bolt that goes throuth the short shaft to set belt tension. Once tension is set, the mounting bolts are torqued.
The pulleys will be 3.5" wide blower pulleys and an 85mm wide belt for best wear and pulse absorbtion. I need a 1.5 ratio for my 68" Ivoprop, so I will be using 8mm pitch 60T and 40T pulleys. I'm trying to design the PRSU so the pulleys will need little or no modification.

One thing I want to have done to the stub axle is to cross drill it and drill the end to the cross drill hole and put a grease fitting for easier grease maintance.
Anyone see any problems with having the axle drilled.

ZZ
If you don't have machining equipment you could consider the option of belt reduction systems like those for custom and drag race motorcycles.

BanditBeltDrive.jpg

The system is solid for more horsepower than your engine so it could be reliable for you if an aircraft engineer approves the concept in your airplane. The ratio could be chosen to the one you mentioned and there's no need for pulleys. But it would be wise to use dampers at least on one side, belts are not that flexible as people think, an O-ring chain drive with sprockets could have more flexibility than a belt drive. Bear in mind that there will be some power lost with the reduction and you might not have all the mathematically calculated horsepower at the prop axle.
 

DangerZone

Well-Known Member
HBA Supporter
Joined
Sep 5, 2011
Messages
2,110
Location
Zagreb HR
Interesting subject, especially as I am considering dampers on my conta-rotation drive system. View attachment 20549 This early model shows a belt for the primary drive but I went away from that idea due to my fears about bending belts backwards. Even so, a chain drive still needs to isolate engine pulses..or does it? Will my 3 gears take out the shocks? If not, then 2 idlers of different rating is my solution. A strong one on the drive side, and a weaker one to take up any slack due to chain stretch (from heat) etc. Yes, I am aware of the extra weight. I realise that my initial drive gear should be smaller for more torque.
Wouldn't such a system have serious safety issues if these chains achieve high speeds..? The slack on both sides could drop the chains off the sprockets if they are longer, even with guide rails the problems of mass at 3G turns would flip the chains off or twist in all possible directions.

Did you consider the option of guided shafts? They could be a bit heavier if properly closed and sealed but they would provide more safety and solve the problem of lubrication that chains need.
 

Dan Thomas

Well-Known Member
Joined
Sep 17, 2008
Messages
5,230
Dan,
So the transmission effectively isolates the main and tail rotors so no resonant frequency can develop? I knew this was comparing apples and bannanas, but was worth throwing out there. i still think it might be worth pursueing, just for fun. Also, the idler does not have to be very heavy, and I would think that the spring tension would keep it from bouncing enough to let any real slack in the belts. I would think it would be best to use just one idler in a v belt drive, grooved pulley pulling out on the belt, on the lee side of the load, so to speek. This might work well with a single wide micro v belt also, as they tend to be real sensitive to belt tension, IME.

How much time do you have with a belt drive? Was it yours, or a freinds that you flew? I've heard so many different oppinions about belt drive reliability, I am hoping to get a little clearer picture from folks who have used them.

Thanks for your input, This is a great Forum!
Zach
The helicopter transmission has a sprague clutch, for one thing, that prevents the rotor system from driving the engine as the crankshaft slows between pulses. The engine's flywheel is fairly heavy so that it will run without being clutched to the transmission.

I have maybe 20 hours behind the RAF redrive, on a Subaru in a Glastar. It had some issues, too. Around 1400 engine RPM it made some ugly vibrations and noises, less so at 2800. Resonant frequencies at work.

Dan
 

Head in the clouds

Well-Known Member
Joined
Mar 11, 2012
Messages
1,983
Location
Gold Coast, East Coast of Australia
Has anyone had any experience with using silent chain for a PSRU? Silent chain - Tsubaki Europe

I've been toying with it for years and although it's a bit heavy it might solve a few issues. It's way better than standard sprocket chain because it grips the gears as it rotates so there is no relative motion between chain teeth and gear teeth. It's guided by the central tooth so it doesn't have alignment issues, in the area which is not in contact with the gear teeth it has slack and so should absorb torsional pulsing quite happily, and because it isn't run tight like a toothed belt it should be much kinder in terms of radial loads on output bearings.





Ramsaychain website has lots more info.
 

deskpilot

Well-Known Member
Joined
Jun 30, 2009
Messages
1,121
Location
Morphett Vale, South Australia. Just south of Adel
Wouldn't such a system have serious safety issues if these chains achieve high speeds..? The slack on both sides could drop the chains off the sprockets if they are longer, even with guide rails the problems of mass at 3G turns would flip the chains off or twist in all possible directions.

Did you consider the option of guided shafts? They could be a bit heavier if properly closed and sealed but they would provide more safety and solve the problem of lubrication that chains need.
DZ, I went for chain for simplicity and weight. Yes, there are potential problems but then, what system doesn't have. I envisage that the primary chain will be in tension at all times. On the drive side, the heavy spring will hold the chain about 1/2" offset and be strong enough to dampen the engine pulses. The return side will be a normal tensioner, taking up slack from chain stretch only. Not modelled yet, are possible tensioners in the secondary drives. Hadn't though about it till now, but there might be some growth in those chains as well. I think my biggest problem will be engine vibrations a lower speed ranges.
 

deskpilot

Well-Known Member
Joined
Jun 30, 2009
Messages
1,121
Location
Morphett Vale, South Australia. Just south of Adel
Has anyone had any experience with using silent chain for a PSRU? Silent chain - Tsubaki Europe
G'day mate. I looked into these some years ago and remember that they would not sell their chains for aircraft use. Too far back to remember the details and that comment came from the parent company in the states. Going by your image, maybe that's no longer the case.
 

Head in the clouds

Well-Known Member
Joined
Mar 11, 2012
Messages
1,983
Location
Gold Coast, East Coast of Australia
I'm going to be making a cog belt PSRU ......

Before I get one I wanted to know if higher angle thrust bearing would fit, .......

One thing I want to have done to the stub axle is to cross drill it and drill the end to the cross drill hole and put a grease fitting for easier grease maintance.
Just a comment about PSRUs and their bearings learned from hard gained practical experience and many early failures.

In the 1980s we made lots of PSRUs because they weren't commercially available and we were using small industrial engines on our ultralights. We mainly used B section V belts and if the system was to last and/or be reliable for more than about a hundred hours then the crankshaft pulley needed to be carbon steel and heat treated. The main cause of problems was the rapid wear of the primary pulley. We usually used two belts and they had to be bought as a matched pair. The airscrew pulley was 6061T6 or 7075T651 aly.

We never used tensioners but did have a damping roller on the lazy side of the belts, the later ones were hydraulic from a small car engine's timing belt system, we made a small reservoir to give them an oil supply. Belt tension presetting was achieved by having the propshaft/pulley on an eccentric mount with locking bolts as mentioned on a previous post. The whole unit was mounted to a 20mm/ 3/4" thick aly plasma cut plate.

Final drive bearings on the propshaft pulley caused heaps of problems initially. We decided that 'wheel bearing' style of 'Timken taper rollers' were the very worst decision, they just didn't work with the high gyroscopic loads imparted by airscrews. If you tried to use larger ones then they weren't suited for the rpm and broke down because of that. Most folks then used plain old deep groove bearings and at the horsepower we were using they worked quite well. A toolmaker friend had far more experience with bearing applications and solved the problem completely. He installed angular contact ball bearings back to back with a facility for accurately adjusting the pre-load on them (shims and sleeve between them), correct pre-load was the critical factor. They then resolved the radial and thrust loads correctly. For higher horsepower applications a third bearing, deep-groove type, could be added at the non-prop end of the pulley to provide further prop pulley stability.

Hence no facility for greasing was required for those bearings as they were supplied pre-greased and sealed.
 

akwrencher

Well-Known Member
HBA Supporter
Joined
Oct 16, 2012
Messages
1,167
Location
Gustavus, AK
Remember, what they don't know, won't hurt them, so when you order, make sure they don't know! :) Seems to be a common thing in the hombuilt world......
G'day mate. I looked into these some years ago and remember that they would not sell their chains for aircraft use. Too far back to remember the details and that comment came from the parent company in the states. Going by your image, maybe that's no longer the case.
 

akwrencher

Well-Known Member
HBA Supporter
Joined
Oct 16, 2012
Messages
1,167
Location
Gustavus, AK
Thanks Dan for the Details. Maybe not the best plan then.

Thanks, HITC, also, for your experience. A few good sources of info can sure save a guy a lot of time and headache in the workshop:)

That silent chain looks pretty neat. Would be fun to play with for sure...
 

DangerZone

Well-Known Member
HBA Supporter
Joined
Sep 5, 2011
Messages
2,110
Location
Zagreb HR
Has anyone had any experience with using silent chain for a PSRU? Silent chain - Tsubaki Europe

I've been toying with it for years and although it's a bit heavy it might solve a few issues. It's way better than standard sprocket chain because it grips the gears as it rotates so there is no relative motion between chain teeth and gear teeth. It's guided by the central tooth so it doesn't have alignment issues, in the area which is not in contact with the gear teeth it has slack and so should absorb torsional pulsing quite happily, and because it isn't run tight like a toothed belt it should be much kinder in terms of radial loads on output bearings.





Ramsaychain website has lots more info.
Could you please explain why would this chain be better than a standard sprocket chain?

From another point of view, this chain is heavier than a simple 530 X-Ring chain for motorcycle use. It produces more friction meaning more heat. It is mostly used inside the engine and it is sensitive to dirt, dust and debris. It needs more lubrication. If your engine turns at more than 3000rpm (and most engines do) you need a high speed silent chain.
On the other hand the X-Ring chain has rings on it, meaning more roll and flexibility and less friction. Less weight. It is cheaper. It's wear can visually be noticed. It is less wide which allows double slacks for more security. It can withstand the crankshaft sprocket to turn at higher engine speeds. It is more elastic. It is proven to be dirt tolerant. It might not need a tension pulley if properly designed meaning less moving parts.

I'm really trying to understand what are the benefits of using a silent chain in reduction units. It isn't the sound because we'd have the prop and engine turning it making more noise than the chain itself. So what would be the rational reasons to use such a chain..?
 
2
Top