Transmission to shaft coupling bolts sheared '07 H38

[Ralph Johnstone]

So possibly over torquing along with misalignment?

If one's bad, two's worse.

Over torquing is not required to ensure the nuts stay put. To make sure you sleep well at night, use the following together on each bolt:

1. nylock nuts

2. Loctite 242 medium strength thread locker



3. a star lock washer.

That is NOT coming apart unless you really want it separated.

 

[Ward H]

My new Twin Disc TMC 40 tranny came with Grade 8 bolts. Original bolts were Grade 5.
Somewhere I saw a note not to use SS but can't find that now.
While searching I did find this preview paragraph but I couldn't drill down and find the whole topic. It's in line with comments from others.

 

[JamesG161]

I assume the loads are pulsing shearing loads

A lot of forces are in Play.
1) Engine in Forward
2) Engine in Reverse
3) Motor Mounts flexibility
4) Cutlass Bearing wear
5) Shaft alignment.

The real question here is "what caused the cyclic tensile stress ?" I got two nickels, a dime, and a ball of pocket lint that say's it's an angular shaft misalignment problem. Only an alignment check will give us a few more "maybes" to work with.

My bet is with Ralph...
Jim...

 

[jssailem]

I got two nickels, a dime, and a ball of pocket lint that say's it's an angular shaft misalignment problem. Only an alignment check will give us a few more "maybes" to work with.

Sure sounds logical Ralph.

If one bolt looks like it backed out, It is possible (my non-empirical hypotheses) that the one became loose before the second one sheared.

 

[flightlead404]

If one's bad, two's worse.

Over torquing is not required to ensure the nuts stay put. To make sure you sleep well at night, use the following together on each bolt:

1. nylock nuts

View attachment 206341

2. Loctite 242 medium strength thread locker



3. a star lock washer.

View attachment 206342

That is NOT coming apart unless you really want it separated.

Interesting. My experience is that mixing additional locking methods (with the exception of safety wire) can often REDUCE the locking ability. e.g. thread locker and nylocks.

The primary concern here is not the nut backing off, but on the bolt snapping. I think I'll stay with one locking method, probably a nylock with a flat washer underneath, and torque to the standard value for that material, shaft, and thread. I'll have to look it up, but M10 I think.

My plan is to 3d print a couple of blocks to fasten to the shaft either side of the flange so I can use a small metal straight edge to check/adjust alignment in two orthogonal planes around the shaft.

The engine mounts have not been touched, but perhaps something with my new cutlass bearing threw off the alignment somehow.

 

[flightlead404]

Sure sounds logical Ralph.

If one bolt looks like it backed out, It is possible (my non-empirical hypotheses) that the one became loose before the second one sheared.

Yeah, don't know the order of operation here. Possibly one backed off and the others were loose, leaving one to take the load which eventually snapped. I will check the torque on the remaining two before I do anything else to see if they are loose.

 

[JamesG161]

Yeah, don't know the order of operation here.

From your owners manual...

Jim...

 

[flightlead404]

View attachment 206441


From your owners manual...

Jim...

Thanks. I meant, order of operations in the sense that bolt broke and the other backed off, or one or more backed off resulting in the other breaking.

Same gap as the valve clearance on my car which I just set this weekend. I have loads of shim stock and feeler gauges that thickness

 

[JamesG161]

Follow @Ralph Johnstone method.
Jim...

 

[flightlead404]

McMaster-Carr has these Grade 10.9 Black Ultra-corrosion resistant components rated to withstand 1,000 hours of salt spray. Sounds like they would work well.

 

[Ralph Johnstone]

My experience is that mixing additional locking methods (with the exception of safety wire) can often REDUCE the locking ability. e.g. thread locker and nylocks.

I was just taking it to the extreme. Wouldn't expect anyone to hang all that junk on a bolt. My four bolts on the mating flange only have a star lock washer and no problems after 24 years.

 

[flightlead404]

I was just taking it to the extreme. Wouldn't expect anyone to hang all that junk on a bolt. My four bolts on the mating flange only have a star lock washer and no problems after 24 years.

Does anyone bother to safety wire things like this on a boat? Its something I do regularly on my other motorized toys.

 

[Ralph Johnstone]

My plan is to 3d print a couple of blocks to fasten to the shaft either side of the flange so I can use a small metal straight edge to check/adjust alignment in two orthogonal planes around the shaft.

Sorry, but that's not the alignment I'm referring to when I say angular alignment. The idea of the small blocks will give you radial (movement away from the axis) alignment only and all you need for that is a bit of touchy feely with your fingers around the flanges. The engine sits on motor mounts and requires next to nothing to move it radially. Close enough. It's the angular alignment that's the fun one (aka absolute bastard) to set up. The method I use is shown below.

You're shooting for 0.001"/inch of flange diameter. My first alignment from the factory was out 0.008" on a 4" flange. It was a long, wet, sloppy winter and so I fooled around with it until I got it to 0.002" error. I check it every 3-4 years (or when I remember) and have only nudged it once in 24 years, several years ago when it was edging towards 0.003". An easy job to do with that small amount of correction.

We all want to know (and well deserve to know) what the initial error was when you check. Let's see who had the larger % influence in the failure ................... over-tightening or misalignment.

 

[ice breaker]

There are lots of bolts that after commissioning and some time in service are supposed to be re torqued by the dealer. NEVER HAPPENS.

On our 2004 H41, we were the second owners. 250 miles on the boat in 5 years. 110 on the gen and motor. Windless almost fell off the boat and all 4 bolts connecting the transmission to the drive shaft were finger tight when I bought it. Manual says that all of that is to be re torqued at first service by the DEALER. Never happens. I fixed it and 13 years and 10,000 miles later everything is still tight.

I would suggest that you go through everything and make sure that any critical bolt/ fixture is torqued correctly. Things need to take a set after manufacture, and the re-torqued is critical if it has not been done.

 

[jssailem]

I fixed it and 13 years and 10,000 miles later everything is still tight.

It is what happens when an owner who cares and knows the value of good seamanship maintains his boat.

 

[flightlead404]

Initial Vertical Misalignment: 0.01825"
Initial Horizontal Misalignment: 0.00425

Final Vertical Misalignment: 0.00075"
Final Horizontal Misalignment: 0.00025"

I have 5 decimal places in the results but measured using .001" increments on the gauges. The "feel" of a feeler gauge is, obviously, rather interpretive. Additionally the measurements themselves are not particularly repeatable. I found that repeatedly measuring, rotating the shaft or engine 360 degrees and remeasuring provided results up to .002" different. All this ends up with results that are "a high degree of precision with dubious accuracy" as my old Dad used to say.

That said, the final result is certainly much better than initial measurements.

I'm a little confused about the section on "out of trueness".

The text reads:

Now that we have determined the misalignment, it is a simple matter to apply those figures, as a correction factor, to our measurement data. Then we can see what is left. Take the two tables from Fig.3 and construct two more corresponding to those, except subtract 8 (the vertical plane misalignment) from each of the numbers in the first two columns, subtract 6 (half the horizontal misalignment) from each number in the second columns, and add 6 (the other half of the horizontal misalignment) to each number in the third columns.

I believe I should be adding or subtracting the absolute value of those misalignment numbers, but the text is unclear.

That is, if my vertical misalignment is -8 I'm still subtracting 8, not subtracting negative 8 or in order words adding 8.

Can anyone please confirm or deny my understanding?

 

[RoyS]

Suspect loose bolts. Once one loose bolt falls out the load on the others increases. Loose bolts produce shock loads across all bolts. Bolts with threads at the flange intersection will tend to propagate a crack. Try to find bolts that have an unthreaded area at the flange intersection. Bolts may have been English size in a metric hole. Elastic nuts are a good idea. I would avoid all types of lock washers.

 

[SEMPERAVANTI41]

I was the first owner of my 07 41DS. I also had the four windlass bolts fall off the studs, which I found laying in the compartment under the windlass. I also had my propeller shaft flange bolts loosen and the coupling disconnect. Locktite solved the problem. I would be leery to rely on the nyloc nut as the transmission will get hot, maybe enough to weaken the nyloc?
I think the sheared bolt problem resulted from the bolts loosening. I don't think a high strength bolt is necessary, but couldn't hurt.