torque spec and cotter pin holes

[Guest]

GR850GZ
Clymer says to torque the front axle nut to 28-38 ft-lbs.

At 28 ft-lbs, the hole for the cotter pin is completely obscured.

At 38 ft-lbs, the hole for the cotter pin is not quite completely uncovered.

No slot went past the pin hole. In other words, there is no intermediate torque that aligns a slot with the hole.

How should this be handled?

 

[gearhead13]

I just guesstimate without a torque wrench. If I dont think it is enough torque at the first alignment, then I turn it to the next alignment.
Alternativly, you could torque it to spec, then drill a new hole in the axle.

 

[SqDancerLynn1]

I would back it up to 28 and put the cotter pin in..

 

[gearhead13]

I would back it up to 28 and put the cotter pin in..

 

[Guest]

If I back it up, it will be at about 20 ft-lbs. If I advance it, it will probably be at about 45 ft-lbs.

 

[Nessism]

It should be 28 ft-lbs minimum. Take it that high and advance it more until the hole lines up.

 

[bwringer]

Consider the function of this fastener -- on a shaftie, it doesn't have to do a damn thing besides keep the wheel from jumping sideways. Extra torque doesn't do anything, so the torque spec is pretty light for this size thread on a steel axle.

In other words, you can do either safely: keep turning it until a slot lines up, or back off to the previous slot.


You could also turn the axle 1/6 or 1/3 of a turn, then try it again. Without heading into the garage to check (hey, it's cold out there) I seem have a vague notion that there are a different number of slots in the nut vs. the holes in the axle.

 

[Guest]

Thanks for the feedback.

I've sort of been thinking along Brian's lines. There was a lot of lateral play until I started tightening the nut.

First, I will replace the washer. After taking it apart again, I see that the washer is scored, which would increase the torque reading due to friction. Why I didn't replace it before, I don't know.

 

[7981GS]

It should be 28 ft-lbs minimum. Take it that high and advance it more until the hole lines up.

This is the correct way to do it.

Daniel

 

[Guest]

The old washer had grooves in it. I bought a replacement, and ground the stamping burr off of it. Put light lithium grease on everything. That removed a bit of roughness. But it still took 60 ft-lbs to get the slot in the castle nut to line up with the cotter pin hole.

 

[7981GS]

The old washer had grooves in it. I bought a replacement, and ground the stamping burr off of it. Put light lithium grease on everything. That removed a bit of roughness. But it still took 60 ft-lbs to get the slot in the castle nut to line up with the cotter pin hole.

You should be fine.
A grade 8.8 M12 x 1.25 bolt has a max torque of 65 Ft.Lbs.
M12 x 1.5 is 63 Ft.Lbs.
As the Axle bolts are probably a grade 10.9, max torque would be about 85 Ft.Lbs.
If the threads are lubricated (depending on the lubricant), 10-20 less Ft.Lbs. less for maximum torque.

Daniel

 

[850GT_Rider]

GR850GZ
Clymer says to torque the front axle nut to 28-38 ft-lbs.

At 28 ft-lbs, the hole for the cotter pin is completely obscured.

At 38 ft-lbs, the hole for the cotter pin is not quite completely uncovered.

No slot went past the pin hole. In other words, there is no intermediate torque that aligns a slot with the hole.

How should this be handled?

Ever had your torque wrench checked for accuracy?

In most big cities, there are companies that do gage calibrations. It can be a bit pricey, but it's not a budget-buster either. I'm thinking of having mine calibrated, just to make sure.....

 

[Guest]

success

success

I haven't calibrated the torque wrench. I'm thinking about jerry rigging a home calibration device.

I started over, because the speedometer drive moved the first time. Since repeating the work was necessary, it was prudent to think through the process. An internet search showed that 60 ft-lbs is indeed acceptable for a 12 mm grade 8 bolt, even with some lubrication. It also showed that some engineers devote their professional lives to tightening fasteners correctly.

Why would castle nut slots not align with the cotter pin hole at the specified torque? A nearly equivalent question is, What would it take to insure that the slots and pins align?

Answer:
1) Controlled, constant elongation of axle as it stretches while being tightened.
2) Excellent control of dimensions of all components that go on axle: spacers, castle nut, washer, wheel, speedometer drive. The total tolerance would have to be within better than 0.001".
3) Cotter pin hole and castle nut slots aligned within a few degrees of the first land of the threads.

That ain't gonna happen.
Implication: What I encountered must be common.

Some math:

The axle has fine thread pitch, 1.25 mm per turn. That is 0.0035 mm per degree of rotation, and 0.104 mm from the cotter pin hole being aligned with a slot, to being blocked on center. The axle must stretch that far to turn the castle nut 30 degrees, if it has been snugged up to begin with.

When I put it together last week, the hole was fully blocked at 26 ft-lbs, and the hole was aligned with the next slot at 60 ft-lbs. That gives a figure of 344 fl-lbs of additional torque needed to elongate the bolt by 1 mm. Sounds like a lot, but where I was able to find data on the internet, it is in the ball park. To increase the torque from 26 ft-lbs to 38 ft-lbs would correspond to axle elongation of 0.037 mm.

I went to the hardware store and bought one of each washer they had in 12 mm, 9/16" and 1/2". I measured the thickness of each washer. They were 1.58, 1.72, 1.85, 2.16, 2.31, 2.44, 2.44, and 2.73 mm thick. The washer I used last week was 2.62 mm thick.

Threading the nut on the axle showed some stiction. It didn't come off with a wire wheel. A previous mechanic probably used thread sealant. Soaking it in brake cleaner, while running the nut up and down about 30 times cleaned the threads..

Using 2.44 mm thick washer aligned the slot and hole adequately at a torque of 35 ft-lbs. It doesn't agree well with my calculations, but now I think that I understand how to deal with this.

 

[850GT_Rider]

ok, now my head hurts.....

But you make some very good points, to which I'll add one more. Unless it was a new axle, it retained some stretch from (all) the previous times it was installed and tightened until the castle nut lined up........

 

[Guest]

I could be mistaken....

If the nut was never tightened past ~125% of the maximum torque for the axle, the axle should not have permanently stretched.

125%, because the maximum torque is set at 80% of the yield strength of the bolt. (100/80 = 1.25) That is the maximum torque for the bolt, not for the application. Another guess here: Suzuki engineers probably decided to write a torque spec low enough that another good turn of the castle nut to align it would not exceed the max torque for the axle. If they would have done it my way - swapping out washers/shims, additional assembly time would have been required, and Suzuki would have needed to maintain an inventory of washers/shims of specified thickness.

Yield strength is defined as the stress at which the bolt just begins to permanently stretch. Below that, the bolt should return to its original length. The bolt is a spring in one dimension. To a good approximation, a spring returns to exactly its original length when the stress is removed.

And it all relies on the above, If the nut was never overtightened. Given what I'm finding about the work of the previous owners, that is a big assumption.

 

[bwringer]

This thread is clear evidence that winter is too damn long around here...


But it is still an excellent idea to give some thought to your threads and fasteners before you just grab a wrench and give it one grunt for little bolts, two grunts for big 'uns.

Over on the V-Strom side, we've found that the official published torque specs are far too high for the lightweight hollow aluminum axles used on the V-Strom. By thoroughly re-thinking the engineering and physics involved, we've arrived at revised torque specs and procedures that are still perfectly safe as far as keeping the wheels attached, yet will help ensure that the aluminum threads don't gall.

What's interesting to me is that there are still lots of V-Strom people who can't bring themselves to distrust the ridiculous number Suzuki published in a book. The mere fact that it's been published is magical evidence that it has to be correct. So they overtorque, and at the next tire change bemoan the cost of replacement axles, or post a frantic plea for help with an axle nut that won't come off.

This is magical, mindless thinking. It's far better to be a mindful mechanic -- develop an understanding of the materials and physical principles you're working with, so that you understand when things make sense and when they don't. Motorcycles are pure physics in action, and every interaction -- even as humble as tightening a nut -- has something to teach us.



As far as the much tougher steel axles used on a GS, the torque values under discussion above aren't even close to the physical limits of the material. So there's a lot of room for errors that won't cause any harm. If the threads are still in decent shape and the axles are straight, don't worry about what some PO did -- the threads would fail long before the axle could deform permanently.

 

[Dogma]

Aw, now you got me thinking...

A threaded fastener connection only needs to be tight enough to make sure the fastener is never stretched to the point that the clamped parts don't lose enough compressive pre-load to allow slipping. On a front axle, there isn't anywhere to slip to, so you just need to make sure there is enough tension in the axle to hold all the parts snug against each other.

Overtorquing, the bearing spacer will probably buckle before the axle breaks. But adding too much tension could conceivably reduce the axle's margin in carrying the weight of the bike. Tension and bending stresses add up in a way analogous to tire friction under braking and cornering. But I doubt this is a real danger for GS axles. According to my old Holo-Krome slide calculator, an M12 tightens to yield somewhere around 170 ft*lb. And then the axle is larger than the threaded portion on the end. This being a safety critical part, there is probably a huge safety factor in the locading calculations. It wouldn't surprise me to learn that the axle is 8x stronger than the worst-case calculations called for.

Undertorquing? What needs to be under tension? The bearing inner races need to be clamped by the axle and various spacers so they don't spin on the axle due to the drag of lubricants and some rolling friction. You can easily turn a healthy bearing with your finger, so this probably isn't significant either. Probably lubricant drag increases with rpm somewhat, maybe mitigated by warming. I'd be very surprised if this turned out to be significant.

The only thing that could cause the components squeezed by the axle to separate is some kind of load that puts the axle in tension. In the case of a front wheel, that'd be the kind of flexing a fork brace addresses causing the various parts to lever against each other. But that same loading case also forces the individual components against each other. So there really isn't anything that can cause an axle with some tension on it to allow parts to slip.

So, the front axle needs to have a some tension, with a lot of margin on the high side. If you're worried about it, I'd say torque to minimum spec, then advance to the next hole. There's probably no problem with backing off from max spec, or going to the nearest hole either, as long as the axle has some tension on it when you're done.