one way needle bearings on a 3x3 trike

S

Sam_Q

1 mW
Joined
May 7, 2008
Messages
13
hi guys first post.

I am custructing a 3 wheel drive recumbent trike and I am thinking of using one way needle bearings instead of this jigger I came up with:

center-coupling2.jpg


This thing acts like a crude differential for my two front wheels, it's basicly a coupling I made that fixes two ratcheting bmx sprockets together. I would use a pair of 25mm I.D bearings on some machined down 1" cro-mo shafts which I would case harden. If this would work not only would it be much simpler and lighter it would also be much easier to package in my very complex and camped frame arangment.

What I want to know off people is how good are these bearings against strong forces? could a 25x32x20 bearing hold the torque of a riders legs after being geared down? I know the question is vauge but I need to start somewhere. I know that these are sometimes used on a motor shaft and I plan to do exactly this for my motor assist but I don't think the force is even remotely comparable because the torque is spread over a high rev range unlike the pedals spun at 30rp/m pushing against a solid seat.
 
Welcome Sam,

You need to know roughly how much torque it will have to handle and then look up the specifications of the one-ways.

A drawn-cup roller clutch or a sprag clutch for that size of shaft should have a nominal rating of about 65 Nm.

Divide the torque at the cranks by the gear ratio in the lowest gear to get the highest torque at the wheel/shaft.

How much torque at the cranks depends on: What length cranks? How heavy you are? :) Probably 100 Nm would be a ball-park figure...
 
the original currie/USPD Prodrive uses a one way needle bearing to couple the output of the 5:1 reduction planetary to the #25 chain sprocket inside the drive unit. the factory ones use a 3/8" roller and fail after a period of use. the aftermarket ones have a shaft adaper and use a 1/2" unit that is much more reliable.

the EV warrior dive unit uses a 1/2" one to couple the 900W total motor power to the roller. it is pressed inside of the drive roller. it is very reliable.

the mid- drive version of the Cyclone uses a 20mm one that acts directly on the planetary output shaft. i've not heard how reliable this is.

a big issue with these one-way drives is that they cannot be used as a bearing at the same time. additional support bearings are needed to keep the driven axle centered within the roller clutch. on the EV-Warrior there are wide bronze sleeves on either side of the roller to make sure the shaft runs true and centered inside of it.

so the question is - how much room do you have for additional support bearings? the rollers themselves should be strong enough for the torque of a reasonably sized motor.

rick
 
thanks for the replies guys,

Now Miles my crank arms are 170mm from center of crank to center pivot of pedal. Weight is irrelevant because it's a recumbent trike and so I have something hard to push against. I would dare say I could do 100kg of pedal force peak for one leg at a time when I am stuck in mud. The gear reduction is 1:2 and I think the bearing is rated at 49Nm. I also think that just as I suspected earlier I would be drasticly overloading the bearing and ripping it to bits.

rkosiorek: thats a very interesting point on the lack of center location, I never knew but it makes complete sense. Now that application you mention thats for the output of the actual motor yes?
 
Sam_Q said:
rkosiorek: thats a very interesting point on the lack of center location, I never knew but it makes complete sense. Now that application you mention thats for the output of the actual motor yes?
Click to expand...
Sam,

You can get roller clutch units with support rollers added (obviously, they're wider). I think I would use a sprag clutch, with a built in radial bearing, though.
 
I forgot to mention I would have lots of room on either side otherwise the sprag clutch would be the go.

Doesn't get around the fact that I want to put probably 5x the rated torque through it. Do you think a simple licking collar on either side of the bearing would surfice?
 
Sam_Q said:
Now Miles my crank arms are 170mm from center of crank to center pivot of pedal. Weight is irrelevant because it's a recumbent trike and so I have something hard to push against. I would dare say I could do 100kg of pedal force peak for one leg at a time when I am stuck in mud. The gear reduction is 1:2 and I think the bearing is rated at 49Nm. I also think that just as I suspected earlier I would be drasticly overloading the bearing and ripping it to bits.
Click to expand...

100kg force converts to 981 Newtons

1M / 0.17 = 5.88

981 / 5.88 = 167 Nm at the crank axle

167 / 2 = 83.5 Nm at the wheel shaft

The nominal rating of the sprag clutches is half the peak capacity - not sure about the roller clutches..
 
Sam_Q said:
I am custructing a 3 wheel drive recumbent trike and I am thinking of using one way needle bearings instead of this jigger I came up with
This thing acts like a crude differential for my two front wheels.
Click to expand...

Hi Sam,

I peeked at your other trike pictures at photobucket, but didn't find the 3wheel drive trike. Do you have any drawings or pics to satisfy the curious?

I take it by selecting gearing you plan to drive front and rear wheels at as close as possible speed/ratios, then when steering, the ratchet/freewheel allows the outer wheel to spin faster, right? You might get a bit of funny steering feeling this way (since the torque generated by the non-symmetric drive will counteract the steering movement. However this may not be a problem with bikes at bike speeds. (It is a problem for some cars with limite slip differentials)
 
Sam_Q said:
thanks for the replies guys,

rkosiorek: thats a very interesting point on the lack of center location, I never knew but it makes complete sense. Now that application you mention thats for the output of the actual motor yes?
Click to expand...

in all 3 cases these handle the output of the motor. the warrior it is only the motor ouput. the Currie and the Cyclone have a reduction gear box between the motor and the roller clutch.

come to think of it, so do many others. the BMC geared hub motor - also has a roller clutch as does the geared Sanyo/Rabbitool and the Panasonic BB drive.

rick
 
thanks Miles, I will have a good look at the calculations and see how you worked it out. However I beleive that due to my lack of a proper explination you made a mistake. The drive is geared so 2 rotations of the crank equals one of the wheel shaft. So would I be right in saying that it would be 167 x 2? so a total of 334Nm?

Now heres another one for you; this would go on just the drive-shaft of one of my three wheels. So wouldn't that mean that the bearing would only see one third of that torque? So 334 /3 = 111.3 Nm. If this is right this tottaly confirms that I will stick with my twin coupling I have pictured above.

This also helps me decide what uni-joints I will use, I will now get one of the uni's I had planned to use and see how much torque it takes to break. I may have to upgrade to a much heavier model.

rkosiorek: I am certain your right about them being able to handle ther motor torque. When I was visiting the recumbent maker Greenspeed the other day I saw they had a prototype electric asist unit, it had the one way clutch on the motor shaft and then on the outside of the needle roller was one ratcheting sprocket to stop the pedals spinning and one normal sprocket. The torque output of all motors is fairly low because it's spread over a high rpm range. However when I am stuck in my and it's me pedaling I might be doing only 2 revs per second with me pushing hard agaist the seat to pedal. If the same power was being delivered by a motor at lets say 3000RPM then it would only need 2% of the total torque. I do hear you about some having the clutch on the output side which would have to take some serious load. I will look more into those that you mention, so thanks for the info.

Jag: thanks for the interest. I have oh so many drawings, unfortunetly not a whole lot else. I will take some photos of my drawings and post the link. Your right on the mark with what I have in mind for the torque split if you can even call it that (probably not). Because of ground clearence issues more than anything else the chain will go from the crank to an intermediate power transfer shaft just infront of my seat. The gear cluster will be on this shaft and will be above the axle line. This will mean that the deraileur will be high off the ground just behind the front wheels line and not asking for a rock to tear it off. From here a chain on one side goes to the front axles with my assembly above and on the other side a chain goes to the rear to turn that. Having 20" wheels all round will mean I will have perfectly even speeds going to each wheel. Now I have riden a tricycle before with only a single outer rear wheel being powered. I found that turning in the direction of the drive wheel would be like having a harder gear all of a sudden because its trying to cover a greater distance than the rest of the bike. For me this is a small compramise as the only real time I would want one wheel to freewheel is when I am riding to where I want to go off-road. Its pretty much to stop my uni-joints smashing apart and my nice chunky tyres wearing out.

I also will have electric assist in the form of an RC motor, but for short bursts only. My build will no doubt take a stupidly long time will be interesting because I look at just about everyone else work and think its all too heavy (but most often significantly better machined). I originally aimed to have my trike running at 17kg (37lb) but now 20kg (44lb) with the 2kw electric assist. I will also have many bash bars over sensative components as I dont plan to treat this thing nicely.

You probably saw it but here my current trike #3

trike-dec-06.jpg


15kg (33lb) last time I checked and apart from the King-pins its all mild steel.
 
Sam_Q said:
thanks Miles, I will have a good look at the calculations and see how you worked it out. However I beleive that due to my lack of a proper explination you made a mistake. The drive is geared so 2 rotations of the crank equals one of the wheel shaft. So would I be right in saying that it would be 167 x 2? so a total of 334Nm?

Now heres another one for you; this would go on just the drive-shaft of one of my three wheels. So wouldn't that mean that the bearing would only see one third of that torque? So 334 /3 = 111.3 Nm. If this is right this tottaly confirms that I will stick with my twin coupling I have pictured above.
Click to expand...

Yikes! That's a great-granny gear...

Correct on the total torque.

Correct on the torque per bearing, if there's equal distribution.
 
15kg (33lb) last time I checked and apart from the King-pins its all mild steel.
Click to expand...

Wow, that's some superb engineering, particularly for something that's designed for offroad.

Can't wait to see what you come up with on the electric drive side. :D

Are you from Melbourne?
 
Miles indeed it is, thats in my lowest gear:

crank 22T to the 32 tooth in the sprocket cluster, then an aditional underdrive with 13T sprockets on the center shaft and 18T sprockets on the either end. So 22:32 and 13:18, using only 20" wheels it means I have a 9.93 gear inches compared to 17.85 with a mountain bike using 26" wheels.

Right now I have a 24:35T combo which is 13.7 gear inches, and even though thats considered seriously low I still get stuck between pedal strokes often when in mud or going over rocks/branches/ridges/etc..


Grinhill: It's some ok engineering but moreso a minimalist approach. For a start:

Rear brakes only- I only have ever had rear brakes on my trikes, this is generally considered a suicide approach but I don't go that fast and I only have bushes and enbankments to run into, not cars. Disk brake calipers and particulally drum brakes are really heavy.

No adjustable boom- this is my trike and its for my size, so that means one less length of material to have the boom slide in and out for different height riders.

No suspension- although this might be odd for an off-road machine most of my riding is tricky low speed stuff/ie: running over everything. Having no moving arms means I can have my seat very close to the frame and chain, so a lower center of gravity. A lower center of gravity lets me have one of the narrowest tracks (front wheel to wheel width) of any trike so I can no only have a super tight turning circle but also fit between trees/posts/etc

I made this thing roughly 8 years ago though, since then I have learned how to do machining and also a heap of other usefull stuff. So this new bike should be interesting. This coupling I have pictured shows how crazy I am going with lightening, every gram I can remove I am removing.

For the electric assist it will be heavily geared down for huge torque at low speeds in short bursts and I plan to have interchangable pinions and an additional bearing support to stop all side load on the motor.

I am in Endeavour Hills, S.E suberbs.

Anyway I have probably given you guys about 10x the info you wanted, hehehe
 
Hmmm! Differential speeds for front wheels on turning a tadpole trike with a motor on each wheel. Can this be achieved electronically, I ask myself? Imagine a scheme where turning the handlebars turns two potentiometers (or some such) and those in turn control the speed of each front motor. What do you think?
 
It's definately possible. However I am sure you would find that one motor would automaticly slow down with the extra load and the oppersite for the other side. I have a design for a "one day" project of a super light mini buggy that powers all 4 wheels with a motor above each wheel as part of the unsprung weight.

Because of the design of my driveline in my trike the electric assist will work for all three wheels at once.
 
Miles, you're such an "Ecyclo-pedia of Knowledge"! Tell us what method you use to store it - bookmarks - database? I suppose you wouldn't consider making a "sticky" thread here that would be a dump of whatever collection you have? :wink: :shock:
Edit: or perhaps in the newly proposed Wiki (Sheldon Brown - eat your heart out!)
 
Your better at Googleing than me then! :lol:
 
Miles said:
http://www.vxb.com/page/bearings/CTGY/One-Way-Bearings
Click to expand...


how would they work, they are not keyed in any way. wouldn't they just slip in the shaft or what ever they are pressed into
thanks
 
You could bond them onto the shaft (Loctite 603). You can also get them with internal and external keyways (not from VXB, though).

The roller clutches don't need anything more than an accurate hardened shaft, of course (and bearing support, if not integral).
 
better late than never, some more basic info:

http://s-86.com/s-trike4.html
 
You must log in or register to reply here.

Similar threads

Reid Welch
"Extreme MY1018" archival, closed out
Replies
11
Views
17,838
GIGATT
G
Back
Top