BMW K1200, K1300, and K1600 Forum banner

Exploring clutch drag......

7K views 22 replies 8 participants last post by  EricK1300GTSE 
#1 ·
The clunk these bikes make when going into first has from day one made me cringe. How can all this good engineering have embraced this clutch? The sounds a bike make are part of the experience. How much do we pay attention to our exhaust note? The clunk is the sound of crashing metal. But this bike is exhilarating to ride. But you have to get past first. Here’s what I’ e been working on.

A few givens here, this is a 2009, K1300R, upgraded oil pump and a K1200 slave cylinder, 32mm. A new given is a pause after disengaging the clutch before dropping into first. No need to pause with any other gear but first. This pause allows the steels to shed oil that otherwise will cause fluid coupling and the clunk. Patience is rewarded. I need to put more time into riding the bike now. There are inconsistencies in how long a pause is needed. I think I’ve added a new bit of personality to this bike. If I find myself in a sentimental moment, I can forgo the pause and the clunk resurfaces as a reminder that the new personality is worthy of embracing.

At this point I will just give an outline of the problems and the solutions I found dealt with clutch issues. For most K owners what I have come up with is beyond what many would want to do or have an interest in. Many of these bikes have made it past 100000 k with no complaints. Also my approach is likely not the the only way. I can certainly give the details if anyone wants to go there. Some may just be interested in understanding the aspects I have looked at in this clutch. I do wish BMW had made the hub maybe a mm taller. That would be a 50% increase in the space for the plates when disengaged. It would have offered the plates more shoulder room and less drag. Oil we can manipulate, hub space is set at the casting plant for the hub.

First there is the issue of oil. Everywhere it has been stated there is not enough oil being delivered to the clutch. I would ask, how did this get established? The early K1200 did have issues with wear and oil. BMW never seems to have put the resources needed into fixing this clutch. They circled around the peripheral, different oil holes, anti-judder setup, anodized basket, steels with dimples, changing the thickness of the steels, the four variations of the thrust adapter, a higher capacity oil pump (really just a change in sprockets), a change in specified oil, these changes were over a number of years. Cynicism whispers in my ear, they needed to give dealers something to show BMW was working on this. Being told ‘They all do this’ is to normalize a defect. I bought into this with the upgrade to the oil pump. It may have been the right thing to do. I have nothing else to compare it to. This is the pump I’ve done the exploration with.

Oil distribution within the clutch is a big problem. What ever pump or thrust adapter the pattern is, to varying degrees, the same. Very oil soaked plates near the hub going to dry plates towards the middle of the pack, then uneven near the pressure plate. Those very oil soaked plates contribute to the clunk. With this bike there is way too much oil being delivered to the clutch. An amount that will always cause fluid coupling.

I filled all the oil holes in the hub. Through trial and error I settled on size of hole and pattern. I think this could be improved, but it’s very time consuming. To give an idea about how much of a change, take the holes at the hub that deliver oil to the hub and anti-judder setup. OEM hub has two holes that are about 3mm, close to an 1/8” in diameter, 180 degrees apart. I settled on one hole that is 1/16”. Sorry my drill set is imperial. Even with that reduction there seems to be plenty of oil at the hub.

The architecture of the hub does not lend itself to even oil distribution. It’s like the designer of the hub and the person who decided where and how large the oil holes were to be, never knew what the other was up to. Just looking at the hub one can make valid assumptions and predictions about the short comings.

Next was to reduce the amount of oil on each steel surface. The goal is a relatively even oil film throughout. This I did first with the oil hole size and multiples and then by grooving the steels. In big commercial clutches grooves are most commonly done within the friction material. Some do groove the steels. I had zero confidence I could do this in the frictions. So grooves in the steels help shed oil. As the friction pads slide over a groove, it acts like a windshield wiper pushing the oil into the channel and draining off the plate through release from the centripetal force.

At this point there was little consistency in overcoming the drag/fluid coupling. I started thinking what would be nice is a two level of oil delivery. A good oil flow when the clutch was engaged and a lesser flow when disengaged. Oil in these clutches preform three functions, lubrication, cooling and flushing wear out of the clutch. I did come to realize this clutch does have two level delivery. The two slots in the thrust adaptor are where oil is delivered. The exposure of these slots is shortened by the amount the slave moves the slotted pin into the transmission shaft. But the effect is not really noticeable. I started putting washers of different thickness on the slotted pin between the transmission shaft and the thrust bearing. Effectively shortening the exposure when engaged and significantly reducing oil flow when disengaged. An improvement, yes, but still erratic.

The clutch plates were not separating consistently. This clutch relies on a completely passive way for the plates to separate, simply, the pressure plate moving away from the clutch pack. I kept the ‘anti-judder’ setup because it provided some active plate separation with the Belleville spring. BMW had at one time introduced those green o-rings to help push the plates apart. They used them on three frictions 180 degrees apart and altering every other friction. I expanded this to five plates with o-rings on successive frictions. Instead of stacking these o-rings as BMW did, I offset these o-rings in a spiral. The spiral shifts the pivot point the friction plate rocks on its O-rings from its neighbours rocking point. Sort of encourages more random jostling of the plates which helps break the fluid coupling'. That’s the idea anyway. By saying offsetting the O-rings, I mean in plan view each successive friction, the tab with the O-ring shifts, in this case counter clockwise by two tabs.

I understand many will be wary of reduced oil in the clutch. My confidence was raised with some reading up on synthetic oils. What they offer is really very impressive. The reliable thickness of the oil film is much reduced compared to conventional oils. Secondly, these oils behave in a non Newtonian way. I had to read up on this. Under heat and pressure synthetic oils become almost solid, might be more like a wax. This produces a very slippery surface that has filled the hollows and valleys of the steels. The anticipation here is wear on clutch will not be increased by a reduced oil film.

Lastly, I am not an engineer or mechanic. I used the same route and stopping points for every test run. I did have to lengthen the run when I realized how long it takes the oil to get up to its running temp. Because I was interested in the disengage state of the clutch, I always held the clutch lever in for about 5 seconds and shut off the bike while disengaged. The clutch pack was on the bench in about 12 minutes.

I will keep an eye on wear. At this time, the hub and pressure plate look to have a clean oil film and more oil than any of the internal steels. At one time in this exploration the oil on the pressure plate was sooty looking, too much pressure while disengaged. I found the pressure plate oil delivery method to be very passive and indirect, a bit like crossing your fingers and hoping . It really has little to do with the way oil is delivered to the rest of the clutch pack.

Finally, this clutch is called a wet clutch. Once I started going in the opposite direction from this clutch needing more oil, I began to consider it more an oil mist clutch. Still classed as wet clutch but oil mist frames the idea of how this clutch works better.


I put a video on YouTube 2 months ago showing a diminished clunk. I was still working towards consistency. It is still audible but not as harsh, I can’t feel it through the bike. The second video, most recent, has almost no sound going into first. I have to partially let the clutch out while in first to show it is actually engaged. The first link is the earlier video. The second link is the most recent one.

https://youtu.be/F_BiZiBCbtA

https://youtu.be/x92YQFi4WUo

For anyone wanting a good overview about the K1200 and K1300 clutch, go to YouTube and look up Patrique Hofman’s technical review of the K1200 2008 clutch plates and basket. He has a number of videos about these bikes and they are all very informative.

https://youtu.be/IdIi3gZq87A
 
See less See more
#2 · (Edited)
Have you seen John Sykes' posts on ibmwr, or his videos? IIRC John claimed to have been involved in factory clutch design efforts as well has having considerable non-OEM experience. His videos are still up, he posted them under Youtube username "sharealike". IMHO his videos were a master class in presentation of his observations, interpretation and reasoning. He has been to this point the most reasonable presentation I've seen. He also made his stuff available as a package to those who took the time to contact him, although I understand it wasn't easy. He too has a life.

The ease with which your K1300S shifts is a major difference from my 2010 K1300GT, though you didn't show how it behaved prior to your modifications. I have to start in first or it'll clunk like a b*astard; I have to be rolling to get back to neutral when approaching a stop, or I end up banging between first and second six or eight times, by which time the traffic light turns green and I have to bang into first again. Supposedly the PO installed a Barnett clutch but I haven't been in to verify this. It appears some kind of clutch work is in my future. The bike has 30,000 miles.

Pat Hofmann's 3/2017 video blames insufficient oil flow from a misplaced first drilling on the clutch housing, plus the "pin" in the shaft (which on later models has two grooves) not allowing sufficient oil flow. John Sykes went further with a recommendation for dimpled steels (like the UJMs that BMW ought to have copied in the first place) and discussed whether the Belville washer was a good idea or not.

What exactly is the summary situation with the mods you've made, and is it available to "the rest of us"?
 
#3 ·
I seen ‘sharealike1 ‘s postings and videos. Like you, I found them very informative. His focus is on the springs in the basket and clutch wear. Do does mention drilling new holes in the hub for oil delivery. But he never put that part on video, that I have come across.

I never learn to take pictures before the renovation, still! I can describe the sound going into first from neutral. It was a loud thwack, metal on metal with an edge to it. Not just two pieces of metal going head to head but with a sharper second tone. The easiest to ‘note’ was the bike would have a small jerk. This is no light weight of a bike. Think of the force needed to make the jerk happen. How big of a rubber hammer would I need to hit this bike and make the equivalent jerk? All this is going on inside the transmission.

A lot of good information, that I’ve seen nowhere else, was lost to us with the closing of I-BMW. Patrique Hofmann’s videos are now a cornerstone of information still available. He covers a lot of ground.

I am happy to share what I’ve done. As I said, it’s been quite involving. This is no quick fix. I have kept track of where I’ve have gotten to.

I just made a video and posted it on YouTube. The visual library that will not disappear. This video looks at the clutch as is, how the architecture works and the oil flow through this design. Looking at where problems are, identifying where to explore. I plan a second video about what I’ve done to address those areas.

https://youtu.be/PIw7lWHE0PY
 
#5 ·
Will do. This is not over yet if ever. I did a short 1400 kilometre trip and nothing looked out of sorts. The oil at the hub and pressure plate was clean, no ‘soot’ (beginnings of the paste) Granted this is a blink of the eye in the life of a clutch. Even so, I took it as a positive that there was nothing glaring to see. I’m still trying to figure out the erratic aspect of no clunk and next time there is a clunk.
 
#6 ·
Clunk or no clunk is simple to explain......

Engine running in "N":
Pull clutch - shift to first - clunk
Thats because the first half of the gearbox is spinning via the sticky oil plates not seperating, now comeing to a complete stop when engaging the second half of stationary gearbox being held in place by the driveshaft/final drive/wheel.

Engine running, pull clutch, come to a stop while riding:
Both parts of the gearbox are still engaged, overcomeing the stickyness of the now free floating clutch discs..... keeping the clutch pulled, shift to neutral, plates are still floating, now back to first gear. No clunk.
 
#7 ·
Patrique, the procedure you are describing is what I use to check on the clutch drag. Bikes may vary, so I’m describing how mine behaves. I do get it that the source of the clunk is the gearbox. And the cause of that is drag within the clutch. Someone mentioned having success with rolling the bike forward or backwards about a foot before dropping into first. This eliminated his clunk but not for my bike.

To check to see if a change I had done, oil holes and or plates, made any difference, I use neutral into first back into neutral and then back into first. All of this with the clutch continuously disengaged. The real tell for me was how long could I stay in the second neutral and drop into first with little or no clunk. Again this is all done with the clutch lever against the grip. Initially drag spun the plates up almost instantly resulting in the clunk. This is my measuring stick, how long till drag spins the plates back up. I’m not capable of discerning changes in the sound of the clunk. But the amount of time before drag spins the clutch up again, indicates if alterations are having an effect on the drag. Now often the plates are stationary. The alignment in the gearbox has stopped between engagement points. I have to let the clutch out to move the gears (or is it forks?) until I hear that click, which is well before the bike would move.

The other measuring stick is how long do I have to hold the clutch lever to the grip before dropping into first with little or no clunk. This is from idle into first. As I noted previously, with the changes I’ve made, it seems the steels need to shed oil when disengaged. My goal is to shorten this time as much as possible. At the moment the pause is about three or four seconds. If I try to go straight into first it’s just like the good old days of seeing people’s head turn in my direction. This is just a general description. There are so many variables I’m still fussing with. Like coming off a highway run as opposed to riding in city traffic. I’m suspecting the oils viscosity varies enough to affect the clutch, cooler on the highway compared to the stop and go of city traffic. The oil cooler is passive and only effective when the bike is moving.
 
#8 ·
Patrique: I'm in agreement about clutch plate drag causing the clunk. In my case every shift into or out of any gear will clunk, most heavily from neutral to first, but also downshifting from one gear to another. A second shift is just as bad as the one preceding, no matter if up or down a gear, or the time between. It hurts my mechanical sympathy to hear the clunk, I'm thinking of what it means for the driveshaft, and every other part in the chain, and wondering what the weak link will turn out to be.

The simplest explanation is that my (supposedly Barnett) clutch is not releasing, or "worn out", but I haven't yet bought or borrowed the tools to get in there. Youtube provides, as does the rep-rom, the procedural info to disassemble the clutch, and the warning that the rivets are tapered and straight shank "normal" rivets should not be used. It's also news that the K12 slave cylinder has more throw than the K13 unit, and of course larger diameter clutch slaves will have less throw. In the Oberon case, 13% less throw, still isn't much difference.

So I'm waiting, waiting, for either of you or a third party to come up with a solution that costs less than the factory lamellar pack. I'm not alone, I don't think!
 
#9 ·
I might shrug my shoulders about a 13% reduction in the space allotment between plates. The K12 slave offers about .125mm between plates. A 13% reduction of .125 is .109mm. This may make no difference to the clutch drag. But these are paper thin spaces where small changes are more likely to have an effect than a larger spaces would. There is a relationship between viscosity, the space between two surfaces and fluid coupling. Higher viscosity can support fluid coupling in a larger space than low viscosity. I don’t know what the viscosity at operating temperature is. But the space between plates provided by the slave needs to be beyond the coupling ability of the oil. So as insignificant as 13% might seem, I want as much in my favor as possible. As things stand with the OEM clutch, this alone does not overcome the drag = clunk.
 
#10 · (Edited)
Scarecrow:

As it happens I am a licensed civil engineer. It's been a long time since my fluid mechanics class in the 1970s, and since I don't use much of that stuff daily, it's a little rusty and to make any useful calculation I'd have to go back to the books. But I'll bet that at the operating temperature of the clutch and the viscosity of oil at that temperature, the shear reduction caused by aftermarket slave cylinder with only minor difference in plate separation is VERY small. To those who swear the Oberon slave made a huge difference I put forth there's something else going on. Once someone's been into the clutch, not everything that makes a difference can be know. But not having gotten in there, I couldn't speculate what. (Well I could, but without evidence it would be useless to propose). As with you I think a plus or minus 13% separation change is negligible.

Better oiling is, as John Sykes proposed, part of the answer and BMW did that among other thing during the "customer engineering" phase. Production tolerance variations could be significant. Better oil distribution may be all that's needed for some clutches; for others not.

Any proposition that LESS oil may achieve the end is interesting and I wait for explanation thereto. Certainly there's a range of "enough but not too much" but we don't have the sample size to go there. My money's on increasing plate separation, particularly where the last plate meets the hub. So I like the idea of the Belville washer as instituted by BMW who have been of the same opinion. If only it were enough.
 
#11 ·
Honolulu

I agree with you, ‘Production tolerance variations could be significant.’ A collection of tolerances on one side of the range could be expressed one way on this bike and another bikes ‘DNA’ not expressed at all. It helps explain the different range of experience members have. There is also personal perception, fingernails on the black board makes me cringe. For another it could be just background noise.

I have kept that bellville washer on the hub. It is an active separation feature. As I think I mentioned somewhere, the difference between the unsprung bellville washer and it’s washer and the friction plate height is very close, I won’t say exact anymore, to the .125mm for each space. When the pressure plate is disengaged, this first plate, at the hub, has to wait its turn to separate. That bellville washer gives a little push to the rest of the plates. I have wondered about a clutch where each pair of plates had a bellville washer. Little dictators, nothing to see move along. I agree more space on the hub for the plates to shift would be where I would start.

I have to think those that designed this clutch had something in mind. The characteristics they were after resulted in some trade offs. This is pure speculation, this bike was designed to be fast through the gears. 2.8 seconds! Is the small amount the clutch lever has to be pulled to initiate a gear change, one of those trade offs? Speed over clunk.

I initially spent weeks trying different ways to add oil. Lack of oil has been the gold standard. I could not change the clunk in any way by trying to increase the oil. This embedded idea of not enough oil I think comes from the wear seen on the hub and pressure plate. I’m starting to think that the ‘clunk’ itself adds to the wear. This rotating mass comes to a sudden stop. Seems there is some sliding of the plates under pressure until they break. One of Patrique Hofmann’s early videos shows the clutch he has removed from his bike. The whole thing was bone dry. The hub and first plate as well as the friction against the pressure plate were worn bare. But all the middle plates looked fine. He remarks that he is surprised it didn’t blow up. There were a few issues with oil nozzle and oil holes. For me this pointed to where the amount of oil was important. The middle of the clutch pack needed just enough oil that it’s shear strength is easily overcome. (Shear strength, that was the word I was looking for. Fluid coupling as a term is more all encompassing. Shear strength is more specific.). At the hub and at the pressure plate is where oil is the most critical.
 
#12 ·
Something of note that has wrinkled my forehead. A ways into the modifications I was trying, I noted that as the oil temperature went up, the clunk went up. I came to understand it takes about 20 minutes of riding to get the oil temperature up to its stable operating temperature. General riding, not city not highway. So why no clunk while heating up but returning when hot? My only thought has been to do with the small oil hole size I’ve settled on. Maybe the higher oil viscosity when not up to temperature can’t get through those small holes in enough volume to create drag. But when the oil is hot it’s almost water like and easily flows throughout the clutch. I don’t really know, but it fooled me into thinking I’d got it! Then when I went for a general ride there was my unwanted companion, the clunk
 
#13 ·
Well, there isn't much that i can add to the clunk problem besides having installed a K1300 driveshaft, which has improved the clunk a little. And having highly modified the clutch pin on a lathe to increase oil flow to the clutch..... my clunk has become acceptable. I also dont have the beville spring, but i do have a new lamellar package on shelf with one in it when necessary.......
I did notice though, that the clunk reduced significantly after replacing the caked/pasted up first clutch plate from my video.
 
#15 ·
Sorry I lost track of this thread somehow.

On point to your comments about clutch plate separation, I understand that the K12 slave cylinder provides more throw than does the K13 slave, does anyone know the actual throw difference?

The most recent Sykes video I've seen shows a contact point worn in by the ball bearing or spherical contact point of an aftermarket slave cylinder (brand not named).

https://youtu.be/9Jg-QH_V5Fc
 
#16 ·
I just read the conversations you are having and it goes wAAAy beyond my understanding. I was a carpentry contractor in my professional life and most of my mistakes I could fix with a little bit of caulk. I will however give you the best feedback I can on my new clutch when I get back on the road.
 
#17 ·
Honolulu
I don’t remember where I read about the difference between the two slaves. But his explanation carried credibility to me by his presentation. Maybe a better educated person than myself could figure it out. With a set volume of fluid, the reduction in the diameter of a piston would result in an increase movement of the piston. I’m probably missing something. But intuitively it seems tantalizingly close.
 
#19 ·
I'm starting to think that until someone stumbles upon a cure for FGC. first gear clunk, that there will be no calculated cure. Apparently BMW couldn't do it with all there engineering guys. I'm surprised that one of the slipper clutch manufactures doesn't make a total replacement clutch for these bikes that is a simple swap out. Seems like a clutch from a Japanese big bike that works without a clunk could be modified or remade to work in these bikes. A clutch is a clutch. They do one thing. A good machinist could make one from scratch.
 
#22 ·
Been researching this clutch design in these bikes and learning about their issues. Main issue seems to be oil flow to these clutches, as these K series bikes (1200 and 1300 engines) are a dry sump motor, so not an oil bath clutch like most big bore jap bikes. So even and consistent oil feed to these clutches is paramount. The other issues appear to be the main needle roller bearing and the slave cylinder push rod, the height of the spider gear springs. Mr. Sykes has done a lot of great work in improving oil distribution and adding the dimpled steel plates.

My current plan for my 09 1300GT is:
  • oil change to Jaso ma2 oil to see if this changes clutch behaviour (not likely)
  • strip out clutch and check that mine has the latest basket, Belleville springs, fine roller needle bearing and 2 groove slave cylinder push rod. (Highly likely given the clutch was replaced under warranty once before according to the PO)
  • check spider gear springs, replace if too short
  • check steel plates and friction plates, replace with dimpled set ( from Mr. Sykes if he's still doing this work).
  • replace needle bearing with the INA specified 35/30 roller bearing.
 
#20 ·
So basically no one has ever solved the mystery of the FGC. Not increasing oil or reducing oil. Bottom line, we don't really know any more about why this happens then before we started. I thought that John Sikes had this resolved but since I-BMW is gone I can't re-read what he posted at one time. Did John cure this issue or not?
 
#21 ·
Hello!
I am wondering what size the O-rings are because I have a BMW K1300GT 2009 without them. The bike has rolled
70 000Km and I can't see any wear on the clutch plates not even the first one but every shift comes with metal clunk, I think the O-rings would fix this sound. Would it also help to use 10W-50 full syntetic oil to ease the clunk?
 
#23 ·
Also I understand Mr. Sykes has resolved this by enlarging and adding strategicslly placed additional holes, plus using dimpled steels (to remove stiction). The rattle is caused by incorrectly sized springs and poor quality needle roller bearings. Replacing these with his springs and the INA needle roller appears to cure this.

There is also the risk over over-oiling the clutch, this creates a fluid coupling or torque converter type set up, this is due to the high mechanical ratio from the master to slave cylinder, stack height and clamping loads in the stack. So simply enlarging the oil holes and galleries is not the full answer.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top