Destroyed! A warning to those using freewheel cranks....

boostjuice

10 kW
Joined
Jul 27, 2008
Messages
969
Location
Canberra, Australia
Destroyed_Dicta_Freewheel_1.jpg

Destroyed_Dicta_Freewheel_2.jpg

Destroyed_Dicta_Freewheel_3.jpg


The above is the result of a Dicta 22T BMX Freewheel adapted to a chainring carrier used in a serial-driven freewheel crank setup.

Freewheel was in service for only ~600klm (370miles) before seizing. It seized without warning, feet lost contact with the pedals, one pedal spinning over smacked my leg in the calf muscle, the pedals grip-providing grub screws gouging tears in it, the whole situation nearly threw me off the bike at ~40kph.

170mm freewheel cranks were used. Rider weight ~70kg (154lbs). Max possible static torque applied when standing on a single horizontal crank arm is therefore ~117Nm (86ft-lbs)of torque. Was mostly moderate assist pedaling only. Max shock loading can only be guessed at. however shock loading due to jumping on a crank arm was minimised though care.

I don't know what sort of steel is used in the freewheel's outer tooth/ratchet ring, but it is definitely case hardened and pretty resistant to a Dremel cutoff wheel which i later used to cut into it as a test of hardness. There are parting fractures in the valleys of every ratchet tooth (stress riser points), and eventually several have obviously sheared right off. Part of one ratchet tooth even compression welded itself to the inner driver ring which is what eventually caused final seizure. I'm still surprised at the lack of warning noises/intermittent lockup before complete failure.

Why didn't/don't i use a sickbikeparts/ENO freewheel instead of a cheap-ass Dicta freewheel i hear you ask?......Because i need a freewheel with dual row support bearings to take the angular loading applied by offset mounted chainrings. This is a necessity during non-pedal assist power transfer when power is being passed though the chainring set in a freewheeling state (when unlocked ratchets/pawls offer no structural rigidity). The Better quality freewheels like the sickbikeparts/ENO use a single row cartridge bearing and therefore are unusable in my setup as a single bearing offers negligable angular force support.


So take warning ye all. Especially owners of standard cyclone freewheels (they are also made by Dicta :wink: ). Generally though, i think freewheels just aren't designed for the levels of torque applied at the crank which are usually 2>5 times greater than at the rear wheel where they are supposed to be mounted. Sure Trials bikes use 18T freewheels, but they are geared so low for fast acceleration that i doubt the rider can develop the same loading as when these same freewheels are adapted to large chainrings that allow much taller gearing (higher loading).

Who else has this happened too????? There must be others out there?
 
BUT ,, presumably the FW seized whilst you were NOT pedaling hard (@ 40kph ?) ,..IE in the "ratcheting" mode. ?
so maybe its not a matter of how much torque it can take, but how well it can stand up to the "unloaded" freewheeling action ?
IE- those little ratchet pawls rattling over the teeth constantly.
Any more photos of the other parts. ?
 
Hillhater said:
BUT ,, presumably the FW seized whilst you were NOT pedaling hard (@ 40kph ?) ,..IE in the "ratcheting" mode. ?
so maybe its not a matter of how much torque it can take, but how well it can stand up to the "unloaded" freewheeling action ?
IE- those little ratchet pawls rattling over the teeth constantly.
Any more photos of the other parts. ?

No, i was seated and pedaling moderately when it seized (accelerating up to cruising speed). Repeated torque overload/fatigue is what caused the final breakage, not a sudden overload. No way does the friction of ratcheting pawls wear out the teeth like that, or even remotely in that sort of low mileage.

Destroyed_Dicta_Freewheel_4.jpg


The pawls are obviously made of harder steel than the ratchet teeth as they are only marginally deformed.
 
Yikes! I started a thread back in Sep asking about the nature of freewheel failures http://endless-sphere.com/forums/viewtopic.php?f=3&t=21181&hilit=Freewheel+failure.

I think I need to isolate the wheel and power source from my cranks by 2 freewheels in series. Any possibility of a freewheel seizing and instantly making any of my bikes a fixie is unacceptable. I've been wanting to go with a jackshaft for a crazy gearing range anyway, and this is a solid reason force that change.

Glad it wasn't a lot more serious Boostjuice. If it happened to me at 40 or 50mph I think the minimum incident would be severe and easily fatal.

John
 
I've heard of this happening before with a cyclone / elation freewheeling crank that seized and pitched the rider off the bike.

Glad to hear you managed to stay upright!
 
Yes , those pawls have to be tough little critters to do that job !
...were all the pawls still in their correct locations ??.. or was it too scrambled to tell ?
I would guess a pawl came out of location and jammed the mechanism... but no way to know for sure
 
Looks like the pawls are out of alignment.... Was the lock ring unscrewing?

I seem to remember that you tried one of the Tensile freewheels?

An ENO with an additional outboard bearing, perhaps?
 
Hi,
boostjuice said:
Why didn't/don't i use a sickbikeparts/ENO freewheel instead of a cheap-ass Dicta freewheell i hear you ask?......

Because i need a freewheel with dual row support bearings to take the angular loading applied by offset mounted chainrings. This is a necessity during non-pedal assist power transfer when power is being passed though the chainring set in a freewheeling state (when unlocked ratchets/pawls offer no structural rigidity).
I'm happy that you were not seriously hurt :).

You think a sickbikeparts/ENO won't work because you have multiple chainrings or do you think it would fail with a single chainring?

Matt used a Dicta on his recumbent with no reported problems and it looks like "angular loading" caused the problem so the dicta dual bearings didn't do the job. Excessive play?

You might want to try (test - with inspections for condition) a Tensile and if that doesn't work an Eno and see if either is more reliable than the Dicta. Or maybe try upgrading the Dicta or Tensile bearings?
 
Hillhater said:
Yes , those pawls have to be tough little critters to do that job !
...were all the pawls still in their correct locations ??.. or was it too scrambled to tell ?
I would guess a pawl came out of location and jammed the mechanism... but no way to know for sure

The pawls and pawl return springs were positioned just as they are supposed to be when I opened it up. What caused the seizure/lockup was the piece of ratchet tooth that sheared off and was subsequently compression welded as it formed a wedge between the inner driver and another ratchet tooth.

Miles said:
Looks like the pawls are out of alignment.... Was the lock ring unscrewing?

The pawls never seated directly in the centre of the ratchet teeth, but that was not a bi-product of an unscrewing/loose lock ring causing axial misalignment between the inner driver and the outer ring. Just dodgy manufacturing tolerances or unopimized/lazy design by Dicta I guess, but I don't think having them centred would have changed the final result though :shock: . Originally the lockring did unscrew almost immediately (even when torqued significantly). I quickly fixed that with locktite and never had the same problem again (I actually had to cut the lock ring out with a Dremel just to open it up after it failed)

Miles said:
I seem to remember that you tried one of the Tensile freewheels?

Yes I've still got that 96 click 18T Tensile. I haven't tried it in a freewheel crank setup yet for two reasons.

Firstly, I never liked the look of the design of my Tensile from a torque handling point of view. It actually looks worse than the Dicta! The problem is the ratchet teeth are very shallow in their protrusion (short ridges/valleys) so as a result the pawls contact almost the very tip of the ridge of each ratchet tooth. This design limitation is obviously a bi-product of trying to cram more ratchet teeth than your average freewheel into a set diameter (chasing greater engagement points). However, even if the steel quality is much higher in the Tensile over the Dicta, it just looks mechanically handicapped and therefore unsafe. I don't trust it..... Maybe the 60 click version aint so bad :?:

Secondly, even with 4 of the 6 pawls removed, the freewheeling drag of the Tensile is unacceptable to me (removing two sets of the triple synchronous pairs means you loose engagement points, not torque capability :idea: ). The drag is the worst of any freewheel I have hand spun. It's also noisy as hell compared to other freewheels, even when packed with grease to try and quiet it down.
The Dicta and Shimano freewheels i own on the other hand are very quiet.
On another note: I found an ENO can also be made near silent by packing the engagement system with grease without sacrificing engagement reliability (An alternative to weakening the pawl return springs like you did Miles). The factory lubes the ratchets/pawls with light oil only.

Miles said:
An ENO with an additional outboard bearing, perhaps?

That was my first thought as well. No way I can see to achieve it though seeing as every set of freewheeling crank arms I have ever played with never has perfectly concentric and true running screw-on freewheeling threads in relation to the BB spindle hole (the only place to mount a secondary bearing). I even emailed trials bike retailers about this issue and was informed that even the most expensive freewheel crank arms have at least some degree of unconformity in these fastening points. There is no great incentive to get these precise as Trials bikes only transfer torque when the freewheel is in a locked state, and low tooth counts (18T) mean chainline wobble is made negligible by small tooth contact radius's.
 
boostjuice said:
I even emailed trials bike retailers about this issue and was informed that even the most expensive freewheel crank arms have at least some degree of unconformity in these fastening points.
I did wonder about that...

[Ed.]
Is it not possible to fix a bearing housing into the neck of the freewheel inner?
 
Dingo2024 said:
this is maybe another freewheel contender?...........although I have no info on it and just recently released
Interesting..... Sealed bearing (singular). How much of it is made of AL 7075, I wonder....? Well, it's light :)
 
What about using a one way needle bearing set up like on the ecospeed reduction unit. The one that comes from the scooter. I saw a crank set up from hightech bikes that was using a vxb 35 mm bearing with no free wheel. It was really quiet. I am not sure how the vxb one way bearing works but there was no "prawns" or how much torque it could handle but it would seem to me it it failed it would just slip and not seize. Just speculation on my part.

I run the Eco speed motor/ reduction unit to my cranks and that little oneway bearing is quiet and has at least 2000 miles of off road use(one years worth). I was not expecting that bearing to handle but have had no problem with it. My cranks are only freewheeling noise. It I could use a one way bearing in the cranks then system would be really quiet!!!

Anyone know of anyone using oneway bearing crank system? I thought elation way working on something!?!??!!
 
waxman123 said:
What about using a one way needle bearing set up like on the ecospeed reduction unit.
You'd need a large sprag clutch bearing, probably a CSK40.......and it still only has one radial bearing...... Torque for weight, ratchet clutches beat friction clutches....
 
recumpence said:
I bet the WI freewheels will last even though they only have one bearing. A number of people are using them with success. :)

Matt

I used to run a WI ENO in my freewheel crank setup. The bearing failed after low mileage. After it failed twice, i gave up on it and moved to the Dicta option. If my twin chain-rings weren't significantly offset from the center-line of the single bearing within the freewheel then it may well have lasted longer. However i can't improve things because of dropout interference due to my unusually large diameter chain-rings and a set gearbox-width which both restrict my chain-line options.
 
Gotcha.

Wow, I am surprised this is the first reference on this forum to this issue. I can completely understand how overhang torque can cause this problem.

I am glad you were not hurt worse when the FW failed.

Matt
 
recumpence said:
Wow, I am surprised this is the first reference on this forum to this issue. I can completely understand how overhang torque can cause this problem.
It has come up a few times before, Matt.

eg:
http://www.endless-sphere.com/forums/viewtopic.php?p=236828#p236828
&
http://www.endless-sphere.com/forums/viewtopic.php?p=309404#p309404
 
Miles and Boostjuice,

I just received a SickBikeParts freewheel crankset with the WI FW, so I'm especially interested in understanding the problem. I understand that Boost's rig forces the freewheel out more to the right than some other installs, but I don't see why that's an issue in and of itself. In the photos here http://www.endless-sphere.com/forums/viewtopic.php?p=309404#p309404 the motor driven sprocket is directly on the FW, so where's the greater leverage compared any motor driven sprocket with a FW?

In the other thread Bobb is having the same problem. Do they both just have a misalignment causing the problem? The only inline pic in the thread linked above looks to me like the drive and driven sprockets are on slightly different planes, but I can't be certain from a pic. That would certainly account for failures. The freewheel isn't carrying the load, it's not freewheeling at high rpms (my cheapie has freewheeled for nearly 10k miles mostly around 600 rpm), and the pedaling torque isn't much different than normal bike use. That tells me that the problem must be a misalignment or something inherently flawed in the overall design of BB FW's.

If it's not a misalignment:
Stoke Monkey's don't have the same issue do they?
Does putting the motor on the inner sprocket solve the issue?
What about 2 FWs like dual bearing support of a shaft?
Maybe a floating sprocket connection at the FW?
Flex causing misalignment under load?
 
Hi,

Matt said:
Wow, I am surprised this is the first reference on this forum to this issue.
Me too, particularly since several people have posted that replacing the cheap Cyclone with an Eno resolved their problems.

Miles:
Miles said:
I seem to remember that you tried one of the Tensile freewheels?
Are the Tensile better quality than the Dicta? If so do you think a Tensile FW might work?

Boostjuice:
Miles said:
An ENO with an additional outboard bearing, perhaps?
Boostjuice said:
Thanks Miles :D that's a clever idea.
I will look into this if i don't dump the whole freewheeling crank thing altogether with a LHD redesign.
Maybe solicit orders and do a small production run?

Or maybe we should contact Sickbikes and see if they will produce it?
 
Hi,
John in CR said:
If it's not a misalignment:

Flex causing misalignment under load?
If freewheeling (no load on the FW) is triggering the problem the cause might be misalignment, caused by flex due to pedaling?
 
I am using Sickbikes WI ENO freewheel crank with a 56T chainring bolted to their 48T chainring. So far no problems, but I am curious what type of setup is having problems.

I would think mine is the least apt to have problems because the motor is the first thing on the chainline pulling directly on the hub. The gearing ratio between the motor and the crank is also set for my relatively low cadence, so the freewheel crank never turns faster than a normal pedaler.

I have seen some builds where the motor pulls on the return chain, so the motor torque is applied directly to the crank freewheel. I suppose this could affect the bearings and the chainring flex, etc.

And of course some builds end up with the motor spinning the freewheel faster than the pedaler can.

Anyway, in my mind, my crank freewheel should last "forever" since I almost always pedal when the motor is running so no pawl action is taking place, and when coasting the IGH freewheel takes over.
 
Back
Top