G oog le BadWeB | Login/out | Topics | Search | Custodians | Register | Edit Profile

Buell Forum » XBoard » Buell XBoard Archives » Archive through September 03, 2008 » Drive belt snapped when coasting and shifting « Previous Next »

  Thread Last Poster Posts Pages Last Post
Archive through August 26, 2008Thruster30 08-26-08  01:35 pm
         

Author Message
Top of pagePrevious messageNext messageBottom of page Link to this message

Pkforbes87
Posted on Tuesday, August 26, 2008 - 06:52 pm:   Edit Post Delete Post View Post/Check IP Print Post    Move Post (Custodian/Admin Only)

Unless your belt broke before fulfilling a reasonable service life I don't think you have a problem. I've never understood the sudden jump in the belt when shifting with the clutch disengaged, but after reading Al's explanation it sounds like even a slipper clutch won't prevent it. Just the nature of the beast.

In my own case, the original belt lasted 8 years and 19k miles so I won't complain. The shock to the belt while shifting must not be rough enough to shorten belt life - it simply finishes off an already aging and worn belt. That's fine with me - I'd rather be confused at a stop light than bounce off the rev limiter due to a broken belt during hard acceleration.
Top of pagePrevious messageNext messageBottom of page Link to this message

Al_lighton
Posted on Tuesday, August 26, 2008 - 11:11 pm:   Edit Post Delete Post View Post/Check IP Print Post    Move Post (Custodian/Admin Only)

When the bike is in neutral at a stop, with the clutch not pulled in, the entire clutch basket is spinning at mainshaft speed. The large spur gear set on the mainshaft and countershaft (the ONLY spur gears in the 06 tranny, all the others are helical cut gears) are mated and spinning together, but the first gear dog ring hasn't coupled that countershaft gear to the countershaft, so the countershaft is motionless, as is the final drive gear set and output pulley.

When you pull in the clutch, you decouple the clutch basket from the mainshaft, but due to fluid coupling, the mainshaft is not likely to come to a rest. It will continue to spin, with some fluid slip, with the clutch basket. When you put it in first gear by engaging the first gear dog ring on the countershaft, this will pretty immediately stop the mainshaft with the typical CLUNK. But if the clutch is properly adjusted, there is no power in the mainshaft, just free-spin due to that fluid coupling. After engaging the gears, the mainshaft comes to a rest and all the slip is between the clutch plates.

HOWEVER, if your clutch isn't completely disengaged when you put it into gear, there is a lot more power behind the mainshaft. So the force of that CLUNK as you drop it into gear is MUCH higher. That's why if your clutch is out of adjustment, or isn't pulled in completely, you either get the louder clunk or the telltale ratcheting of the dog teeth spinning by their corresponding sockets (antidogs? never know what to call those...). With your clutch adjusted, it's always better to drop it in with some authority, otherwise, the ratcheting as you drop it in can cause rounding of the dog end corners over time.

If the mainshaft stopped rotating completely when you disengaged the clutch, you'd have a different problem. If the dogs weren't aligned with their antidogs, you'd probably bend your shift fork as you jumped up and down on your shift lever trying to get it to drop in. So a little clunk isn't such a bad thing.

The force on the belt and such during these stopped bike events is pretty minimal.

Al
Top of pagePrevious messageNext messageBottom of page Link to this message

Retrittion
Posted on Wednesday, August 27, 2008 - 04:51 am:   Edit Post Delete Post View Post/Check IP Print Post    Move Post (Custodian/Admin Only)

I drop into first when coasting down a hill coming to a roundabout (aka circus) in my town as well as some other times and I don't get a "thunk" (and I occasionally leave some rubber down so there is some torque going on) -- however I don't think I ever fully close the throttle, instead I keep the RPMs 2K-ish (I think -- I'll pay more attention next ride) so that might be part of it, and I let my clutch out gently-ish. I don't know that I have EVER had a "thunk" -- would worry me if it happened, personally.

Ride safe!
Top of pagePrevious messageNext messageBottom of page Link to this message

Blake
Posted on Thursday, August 28, 2008 - 07:09 pm:   Edit Post Delete Post View Post/Check IP Print Post    Move Post (Custodian/Admin Only)

"Blake, I believe your recommendation to blip the throttle applies when moving, not when standing still."

It applies equally in both scenarios. The only time blipping is no beneficial is if ground speed is turning the output side at the same rate as the idling engine is turning the input side.

Ever try to push your cold bike in 1st gear on a cold morning by just engaging the clutch? I won't want to budge without significant effort. Why? The thick/viscous lube is very effectively, relative to human effort, linking the multitude of clutch plates amost as if the clutch were disengaged.

When at a standstill, the output side is not turning; it can't as it is connected via drive belt to the rear wheel which is hopefully resting on pavement on not moving. : ) Blipping the throttle in that case, especially for a cold transmission or one that is slightly overfilled with lubricant, will make a huge difference in how much "clunk" the transmission will experience as you put it into 1st gear.

At a stop, the blipping doesn't do a thing to match input and output shaft speeds, but it does fling a bunch of that viscous oil from the clutch so that viscous torque transfer between the input and output is greatly diminished, virtually eliminated.

Once the bike is warmed up, the tranny lube viscosity decreases and thus doesn't cause the same hard viscous torque coupling and thus no clunk.

Still, it's good to blip. : )
Top of pagePrevious messageNext messageBottom of page Link to this message

Gentleman_jon
Posted on Thursday, August 28, 2008 - 08:52 pm:   Edit Post Delete Post View Post/Check IP Print Post    Move Post (Custodian/Admin Only)

Very interesting discussion of what goes on inside the gear box.

However, there is very little discussion of why the belt failed.

I have tried on various occasions to start a meaningful discussion of the mechanism of belt failure, but to no avail.

Surely there are some members of the board who are familiar with the strength of materials.

I for one, flatly reject the notion that merely twisting the belt during installation can cause the belt to fail and I have not seen one single explanation of what the mechanism of such a failure might be.

The tensile strength of the belt is reported to be in the neighborhood of seven thousand pounds, which is similar to that of normal motorcycle chain.

The belt can be twisted with a force that I will estimate to be about five to ten foot pounds or so.

I would like someone to explain to me how such a load, which is three or four orders of magnitude less than the maximum allowable load could have any effect what so ever on the belt.

It is my opinion that the failure mechanism of these belts is as follows:

The belts are made from textile reinforcing strands of Kevlar or Spectra fibers. These belts are manufactured on wide drums and then slit into narrow belts.

This allows the eventual ingression of water as the belts age, depending on a variety of factors. In addition, when the belt is under tension, small cracks may temporarily open, allowing water or other contaminants to affect the cords.

All of the failed belts that I have personally inspected or seen in photos, show what appears to me to be plastic tensile failure of the reinforcing cords.

That is to say, the fibers are extending from both ends of the rupture, showing that the fibers stretched beyond their plastic yield limit, and were permanently elongated.

Normal this occurs when one or two cords have failed first due to weathering, overloading the remaining strands. Inspecting a rupture usually shows that several strands are not fully elongated, indicating that they were weaker in tension than the other strands.

In addition, failure in these fibers can be induced by turning them around a sprocket with a small diameter.
Even though Buell increased the size of the sprockets in the 2004 and subsequent years, they still are rather small, and put stress on the fibers.

Frankly, I am surprised that the liability lawyers at the Motor Company continue to allow the use of these belts on high performance machines like Buells. One of these days someone is going to get killed when their belt snaps, ( I came pretty close myself when my belt failed), and the number of known belt failures would make a case against Buell difficult to defend.

Ropes made from these fibers were used for yacht racing, and I have seen many disastrous failures of these fibers in that sport as well. Only when they are used in applications where they are completely protected from the weather, and in large diameter applications like tires are they 100% safe in my opinion.

In the mean time, I and the rest of the "Chain Gang" will do what all factory Buell race teams do, and what virtually all of the rest of the sport bike world does: run chain because they don't want to take the risk of spontaneous, and unpredictable belt failure.

Thoughts anyone?

Al?
Top of pagePrevious messageNext messageBottom of page Link to this message

Redbuelljunkie
Posted on Thursday, August 28, 2008 - 10:49 pm:   Edit Post Delete Post View Post/Check IP Print Post    Move Post (Custodian/Admin Only)

During my years in the motorcycle industry, I have seen the results of many "spontaneous, and unpredictable" chain failures. After seeing what a chain can do to a bike (or worse- a rider), I would much rather have a belt let go than a chain.
Top of pagePrevious messageNext messageBottom of page Link to this message

Blake
Posted on Friday, August 29, 2008 - 11:54 am:   Edit Post Delete Post View Post/Check IP Print Post    Move Post (Custodian/Admin Only)

Jon,

The 2003 belts were the ones that were so sensitive to being twisted. Twisting would generate torsional stresses which must have caused the layers to delaminate due to the shear between them.

I'd not conclude that exposed protruding fibers indicate plastic failure. I don't think that the fibers are able to fail plasticly; they don't yield like some metals and some polymers/plastics; they tend to fail in a brittle fracture mode. But I'm no expert on kevlar fiber. I can tell you that when a carbon fiber composite is made to fail in tension, you'll see fibers protruding from the fracture surfaces. I do know that carbon fiber composites do not exhibit plastic macro deformation, they fail in brittle fracture with no yielding.

I think you are seeing the same effect on the failed belts, exposed fibers that broke at other than the fracture surfaces, {the exposed surfaces defining the two ends of the failed belt) which were then pulled from the opposite end of the belt and left protruding.

The most likely cause of belt failures, and one often documented, is damage due to debris.

The incidence of belt failure has fallen off by a huge margin with the new belts.

If I'm going cross country on a belt drive motorcycle, I'd pack a spare, suffer the consequences of folding it all up, but rest easy knowing that it will get me to a shop for a nice new replacement. When replaced, the folded-up spare travel belt would go back in the kit.

Or I might pack some 1 minute epoxy and an appropriately sized strip of heavy kevlar cloth. Chances are I could repair the belt on the bike and ride it very gingerly/carefully to the nearest dealership for proper replacement.

I'm kind of surprised that no one has tried the epoxy and cloth patch repair approach.

(Message edited by Blake on August 29, 2008)
Top of pagePrevious messageNext messageBottom of page Link to this message

Blake
Posted on Friday, August 29, 2008 - 12:01 pm:   Edit Post Delete Post View Post/Check IP Print Post    Move Post (Custodian/Admin Only)

"Frankly, I am surprised that the liability lawyers at the Motor Company continue to allow the use of these belts on high performance machines like Buells. One of these days someone is going to get killed when their belt snaps, ( I came pretty close myself when my belt failed), and the number of known belt failures would make a case against Buell difficult to defend."

I think the case would be easy to defend. Chains also fail. I've personally seen more chains fail than belts.

As to the risk of injury resulting from a belt or chain failure, tire blow-outs are much more dangerous.
Top of pagePrevious messageNext messageBottom of page Link to this message

Thruster
Posted on Friday, August 29, 2008 - 05:04 pm:   Edit Post Delete Post View Post/Check IP Print Post    Move Post (Custodian/Admin Only)

Thanks Blake, I stand corrected. Prior to getting underway from a standstill, blipping the throttle while still in neutral, to fling oil off of the clutch plates, can indeed help lessen the fluid coupling between the two sides of the disengaged clutch. I would, however, like to add a little more detail: I think the blip should be done with the clutch lever out (clutch engaged). This helps spin up the mainshaft coming into the transmission. Then, before letting the engine return to idle speed, pull in the clutch lever. Both sides of the clutch will still be spinning rapidly, so the flinging will be more thorough because it will involve not only the clutch plates on the engine side but also the opposing clutch plates on the mainshaft side. Then, continue holding the clutch lever pulled-in for a little while longer. This will allow the freely-spinning mainshaft to slow down due to friction and also due to the viscosity of the fluid bathing the clutch plates connected to it. Because of the prior flinging, the mainshaft will eventually end up spinning more slowly than if there were unflung oil still coupling the idling engine side of the clutch with the mainshaft side (with the clutch disengaged). The objective is to get the mainshaft spinning as slowly as possible prior to engaging it to the motionless transmission (and the motionless belt, rear wheel, and pavement, all firmly coupled together). That is what will minimize the "clunk" when shifting from neutral to first while at a standstill.

One way to easily and reliably (and safely) break a perfectly good brand-new belt would be to do the opposite of what I described in the preceding paragraph. While at a standstill, in neutral, and the clutch lever out (clutch engaged), rev the engine out to 5 or 6 K, and before backing off the throttle, pull the clutch lever in all the way to fully disengage the clutch, and then quickly engage first gear while the mainshaft is still rapidly spinning from its own inertia. The kinetic energy of the freely-spinning mainshaft will suddenly be delivered as an accelerating impulse upon the engaged gear and on through the transmission and to the belt, much like a column of touching billiard balls. Due to the inelastic nature of the belt fibers being held taut, the belt will then instantly break as surely as a light hammer tap shatters glass. This will happen with the clutch fully disengaged and without the bike even lurching forward the slightest bit, since the resting inertia of the relatively-massive rear wheel provides more than enough resistance to ensure that the impulse acts upon the brittle (in tension) belt fibers, not anywhere else. The impulse is so quick that the belt snaps before the bike gets a chance to dissipate the energy in some other way, such as by lurching forward or by chirping the rear tire against the pavement (both of these actions would be too slow to dissipate the energy within the required time to save the belt). As I mentioned in an earlier note, I'm most emphatically NOT recommending that anyone actually try this with his or her own bike. This is purely a thought exercise, unless you're testing bikes for Buell.
Top of pagePrevious messageNext messageBottom of page Link to this message

Thruster
Posted on Friday, August 29, 2008 - 05:09 pm:   Edit Post Delete Post View Post/Check IP Print Post    Move Post (Custodian/Admin Only)

Great inputs from everyone; thanks. I'll just post a copy of my last email exchange with Al about this, then move on. This has certainly been a dead horsebeating times ten! Thanks for all the stamina, for all you guys who hung in there with me.
Top of pagePrevious messageNext messageBottom of page Link to this message

Thruster
Posted on Friday, August 29, 2008 - 06:47 pm:   Edit Post Delete Post View Post/Check IP Print Post    Move Post (Custodian/Admin Only)

I am a bit of a kite-flying hobbyist and use Spectra lines. The stuff is amazing, and very counter-intuitive and unexpected in how it behaves. It is soft to the touch and flexible, allowing knots to be tied in it without giving up strength, yet it is effectively hard and brittle in a tensile direction, as it is almost completely unyielding and inelastic when held taut. Even a hundred feet or so of the stuff barely stretches, enabling very precise and rapid control of a steerable kite. Spectra has also begun to be used for fishing lines in recent years. I'm not a fisherman, but I guess it provides much more detailed feel of exactly what the fish is doing with the worm, if you're into that sort of thing. Anyway, the stuff cannot be cut with a pair of scissors, a knife, or even with wire cutters, unless one has the patientce to really gnash and mash upon it for a while, and the result is then less than pretty anyway; a red-hot knife is required to melt through it cleanly.

Here're my last emails with Al about this whole belt thing:

Me (Oleh, aka "Thruster"):

Much thanks for your detailed explanations regarding my snapped drive
belt and how the transmission works. I thought your info would be very
helpful to others, so I took the liberty of posting your notes on my
BadWeb thread about this issue -- hope that's ok.

If the belt was already damaged or defective prior to the incident, I
guess it's serendipitously good to have done it in at 25 mph two blocks
from home rather than at 85 on the interstate.

I now understand much better how to handle the controls so as not to
cause the "clunk". In addition to your recommendation to blip, an
alternative approach in the scenario I described would have been to
delay shifting until I was completely stopped at the intersection at the
bottom of the hill, pull the clutch lever in, hold it in for a little
while to allow the mainshaft to come to rest (to match the
already-resting countershaft), shift into first, crack a quick smile
(because now it won't clunk!), then let the clutch out slowly to get
underway, as usual.

Al (Lighton, of AmericanSportBike.com) responded:

Its fine. Yea, that is safest but not always possible, which is why I
generally don't coast in neutral.

Me:

Thanks again for your detailed explanations of how the transmission works. As always, that was way above and beyond the call of duty, very helpful, and very much appreciated. In this world of mostly impulsive and impatient people, it is a blessing to encounter a thoughtful and caring person such as you.

The bike finally got fixed and I picked it up this morning. Of course, I immediately applied the principles I learned from the lively discussion about my belt-breaking incident. I can now get underway after warming up without any clunks, and I even coasted in neutral on purpose numerous times and "recovered" while still in motion, also with no clunking whatsoever.

I ran into the salesman who originally sold me the bike last year. As we were discussing the incident, one of the mechanics happened to be walking by, overheard us, joined the conversation, and confidently confirmed the diagnosis. Note that it is not necessary for the belt to have been defective or damaged, contrary to what you and others were supposing. The synthetic fibers in the belt are likely made from an aramid of some sort, like Spectra. One property of such fibers is that they don't stretch at all (practically speaking). If held taut and then a sudden stretching impulse is applied, the fibers have nowhere to go, so they snap, literally like glass breaking when struck quickly, even with a very light hammer. The inertia of the rotating rear wheel thus served as the immovable brick wall that resisted the sudden tensile impulse applied to the belt as a result of engaging, through hard gear linkages, the inertia of the resting mainshaft and clutch basket.

One design solution would be to spring-load the idler pulley. The spring should be very stiff, so that under normal operating conditions, the spring would not compress at all, and the bike would behave exactly as it does now without a spring-loaded idler. However, if any load approaching the tensile limit of the belt were to be applied to the belt, even over a very short period of time as in my incident, the spring would compress and act as a buffer to quickly take up almost all the energy of the impulse, then give it back over a relatively longer period of time as the spring re-expands, enough to effect a more gradual change in the rotational velocity of the rear wheel and perhaps also chirp the tire as well, which also would dissipate some of the energy.

My suggestion for such a fix would not really be needed for me personally now that I understand how to reliably avoid the problem. Nevertheless, in my search for understanding this problem better, I found numerous other people who were utterly perplexed as to why their belt broke. You may have noticed that one contributor to my BadWeb thread even described how he was totally unaware when the belt broke while shifting into first as he approached an intersection. After braking to a stop, his bike didn't move when he tried to get underway again, and when he looked back, he saw the broken belt lying on the ground behind him! I know you'll be tempted again to say "probably defective", but I really don't think so. Those belts are wonderful, but they really do behave like glass under just the right set of circumstances and conditions.

I described the whole story to my best friend, who is a sports car driver and tinkerer, and he noted that the ritual you described about engaging the clutch and blipping the throttle to spin up the mainshaft before disengaging the clutch and shifting into gear (either first or second) is exactly what truck drivers do when they "double-clutch". Because they have so many gears, it is impractical to build their transmissions with synchronizers (and they'd quickly wear out anyway), so when the truck drivers shift, they need to disengage the clutch, shift into neutral, engage the clutch, blip the throttle (to bring the mainshaft to the proper speed), disengage the clutch, engage the desired gear, and finally, engage the clutch and drive onward.

Have a great holiday weekend!

Al:

Trojan sells a spring loaded idler. I have not been convinced of it's
merit, for a couple reasons.

1) The spring looks too wimpy.

2) Any time you see a spring, the first question that you MUST ask is
"Where is the damping?". A spring merely stores energy, it does not
absorb it (with the exception of a trivial amount of hysteresis in the
spring). That means that any energy put into the spring is going to be
put back into the belt. I've spend enough time watching that drum
taught belt dance around like a hula dancer on speed while tuning the
bike on a dyno. You'd be amazed how much deflection that the belt has
while the system is spinning, which is surprising since it seemingly
doesn't flex at all when still. Adding a spring to that doesn't make a
lot of sense considering that.

Me:

The Trojan spring-loaded pulley you describe sounds like it does not meet the requirements that I described earlier, and I'd guess its design is probably not based on sound principles, i.e., it's not at all clear to me what purpose it is meant to accomplish or what problem it solves.

As I described it, the spring mechanism I proposed would not flex at all during normal operation, including during dyno exercises. Its sole purpose would be to buffer impulsive loads that approach or exceed the breaking point of the belt by absorbing the impulse, then delivering the energy back to the belt over a relatively longer period of time compared to the duration of the original impulse. A damper would be fancy but probably not necessary for the stated design goal. The spring would not instantly redeliver the same impulse right back to the belt because the spring would remain flexed as long as its pressure against the belt corresponds to a belt tension near or exceeding the belt's breaking strength (again, not a normal operating condition, but one that seems to occur now and then to unsuspecting riders). My spring-loaded pulley would only be able to break the belt if another impulse is applied at the very instant the spring is bottomed out, though it's practically impossible for that to happen given the mechanisms involved.
Top of pagePrevious messageNext messageBottom of page Link to this message

Blake
Posted on Friday, August 29, 2008 - 08:49 pm:   Edit Post Delete Post View Post/Check IP Print Post    Move Post (Custodian/Admin Only)

"I think the blip (prior to shifting into 1st gear, trans. in neutral with bike at standstill) should be done with the clutch lever out (clutch engaged)."

I agree. Good call and thanks for the clarification.
Top of pagePrevious messageNext messageBottom of page Link to this message

Thruster
Posted on Thursday, February 10, 2011 - 01:27 pm:   Edit Post Delete Post View Post/Check IP Print Post    Move Post (Custodian/Admin Only)

Thanks for your well-considered thoughtfulness, Blake.

This winter, I've added another ritual for getting my Uly underway on single-digit-cold mornings.

One morning earlier this winter, while hurrying to a meeting at work, I disregarded the hard-learned lessons above and snapped my belt while standing still in my driveway. I didn't allow for any warmup time, and the belt just snapped right off the bike when I engaged first. Since I now always have a spare new belt, the repair was quick.

After some thought, my revised startup ritual, especially on very cold mornings, is now to start and warm up the bike in neutral, then shut off the engine, engage first while nothing at all is turning (rock the bike fore and aft a little to make sure it's all the way in), then get on the bike, disengage the clutch (pull the lever in), restart the engine (while first gear is already engaged), and I'm good to go.

The purpose of being in neutral during the first engine start is to minimize the load on the starter (as well as allow me to do other things while it warms up). The purpose of the warm-up is to minimize the load on the starter when performing the second start (with first gear already engaged and clutch disengaged). The purpose of the second startup with first gear already engaged then completely avoids the impulse to the belt (clunk) that normally occurs when shifting from neutral to first while the engine is running.

During the ride, I never use neutral anymore -- I just hold the clutch lever in at stoplights and in standstill traffic. My rides now are completely clunkless, start to finish. Evidence, experience, and spirited sharing among friends reigns supreme. No more magic!

Cheers!
Oleh
« Previous Next »

Add Your Message Here
Post:
Bold text Italics Underline Create a hyperlink Insert a clipart image

Username: Posting Information:
This is a public posting area. Enter your username and password if you have an account. Otherwise, enter your full name as your username and leave the password blank. Your e-mail address is optional.
Password:
E-mail:
Options: Post as "Anonymous" (Valid reason required. Abusers will be exposed. If unsure, ask.)
Enable HTML code in message
Automatically activate URLs in message
Action:

Topics | Last Day | Tree View | Search | User List | Help/Instructions | Rules | Program Credits Administration