Yamaha R1 Forum: YZF-R1 Forums banner

1 - 2 of 2 Posts

64 degrees?? Pfff. I can do 90
103 Posts
Discussion Starter #1
Hey Everyone,

A bit of an odd question. As some may remember, I've been battling an issue where the crank case idler shaft has worn a crack in the boss:



I'm trying to figure out why.

The assembly feels tight with all the parts in place, but you can see from this video there is a fair amount of vibration coming from the starter shaft. Ultimately this may be causing the issue. The question is, is this abnormal? Anyone have a video for comparison?

2015 Yamaha R1 Start Sequence Video

Given that the starter has 2 rear mounting points and the front is simply held in with an o-ring, it seems like you are going to get some vibration:


Perhaps this is intended to dampen the variable load from the start clutch as pistons compress/de-compress. However, the vibration frequency of the shaft seems a bit fast for that. The starter was taken apart and everything seems ok with it.


64 degrees?? Pfff. I can do 90
103 Posts
Discussion Starter #2 (Edited)
I thought I would follow up on this, since I feel like this might be happening to others with the 2CR design.

To back things up, this all started with a small oil leak on the left hand side (white duct tape glue/cotton smeared here makes it easier to see), right above the chain cover/main output shaft. It was at the track, you use what you can to trace it:


Long story short, it's coming from the idler shaft bolt, next to the starter. The copper crush washer is flattened with a tiny gap, allowing a little bit of oil to seep out of the crank case. You can see why in the above post, the idler shaft is has an oscillation. At first I thought it was maybe an out of spec idler gear, which was not the case.

The original starter video from the post above has been reposted here (unchanged):
YouTube - 2015 Yamaha R1 magnet came loose in starter causing vibration

compared to a correct functioning starter after this solution has been implemented:
YouTube - 2015 Yamaha R1 - starter replaced, working correctly

Notice the movement difference in the starter shaft (to be clear, even in the correctly functioning starter IMO there is too much movement still, but more on that later). After re-checking the starter, (I didn't notice this the first time), but we can see what's causing the problems:

Notice the 2 magnets on the left hand side. The glue has let go on one and they are now stuck together. This explains the vibration which has eventually caused cracking of the idler shaft boss after 5000km.

IMO there are actually 5 design flaws here:

1.) The magnets are not slot/key (inset) into the housing. With said design, when the winding assembly is inserted into the starter housing, the magnets would not be able to move out of their slot if the glue lets go. Yes it requires more machining or a different casting, and you would need a bit more material to hold the magnets which adds a few grams. Race teams would pull the starter anyways, so seems like a reasonable trade-off.

2.) There is no bearing in the start motor supporting the shaft. Clearly this is done to save weight, but feels like a shortcut. I discussed this design with a number of mechanical engineers that came to the same conclusion.

3.) The 2x mounting bolts where the starter attaches to the crank case are co-planar but should be wider apart, offset, or closer to the starter spline. The rotational force (moment) is magnified by the much larger distance between the shaft spline and the two bolts vs. the distance between the two bolts. With just a non-structural o-ring to counter it near the starter spline, some degree of vibration is inevitable.

4.) The idler boss on the crankcase is not designed to have any load on it. Given the tiny M6 bolt that holds the idler shaft, a small copper crush washer on the exterior, and an aluminum housing without any sort of sleeving that holds the shaft, these are not loadable components. It looks like there is a small sleeve insert in the idler shaft boss with visual inspection, but there is not. Yamaha has confirmed this.

5.) Yamaha is using glue to hold in place a metal component (each magnet) that has a constant load force applied (each other magnet) 8 times in one component. If any one fails, the component as a whole fails in a non-obvious manner.

What to do about it? Not trusting the adhesion of the other magnets, replace to starter with a 2016 (although technically this could happen again). To mitigate the structural load issue, in particular now that we have a weakened and slightly ovaled boss, first we'll cast a plate out of childrens air dry clay from the craft store, measure, 3D print to test the fit and machine a custom part:


We will also drill out the crack and use green loctite 638 (designed for press fit shafts) which is rated for high temperature oil baths, to hold the shaft in place. This is likely a 1-time deal, and will never come out again. I doubt its possible to get the crankcase hot enough to remove it, but it does allow us to use the entire surface area of the boss to distribute load.

So why not just get it welded? We could, but this means disassembling the crank case and the motor runs great at the moment. Long story short, this is a high pressure ADC-12 aluminum cast, which would require a high frequency TIG weld. Welding cast is tricky due to trapped air, but possible. While we can mitigate the heat/rubber seal issues, Arc'ing across internal bearings is the real concern. Alternatively, brazing would be tough to get the metal hot enough and has its own risks. To be clear, this is not an ideal solution, but seemed the least terrible option.


So instead we drill out the crack and add some 638 in there.


Add the brace plate with a 10.9 x 25mm M6 bolt and a bit of gasket maker and let's hope for the best.
1 - 2 of 2 Posts