OT-looking to gather ideas-how to support cantilevered load on horizontal shaft motor

[rickyb]

And that is what I did. I turned an alignment sleeve (started out as a piece of scrap 2 inch black pipe. ID was a loose slip fit, approx 0.060 wall, a slot the entire length of the part and a second slot opposite the first with a couple of narrow bridges remaining to hold the two halves together. Once aligned I spread the sleeve apart to remove.

Some images attached. The welding of the pump mounting bracket is a bit of an embarrassment - I'm not a good welder to begin with but I used to be able to make a better looking weld with a stick welder. I'm not to concerned about the weld failing. Seems to work well after the first hour of use.

rickyb:
"I see a very typical capscrew fatigue failure caused by inadequate design. The problem is fasteners that are too short, are loosing tension then are fatigued to failure. Also the housing needs to be spotfaced under the screw head and a thick hardened washer added. Rather than a washer I would recommend a hardened spacer about 1” long and 1.25” in diameter to make up for the lack of stretch length in the capscrew. That will fix the problem...."

I gave this post a bit of thought. I agree that using a spacer would allow for proper bolt stretch. I still think the failure is the cantilevered load under vibration. I'm not a Tecumseh fanboy but I seem to have two other similar sized Tecumseh engines and the fasteners that mount the engines to the other pieces of equipment simply fasten thru a 3/16 or 1/4 inch plate into the engine without the need of a spacer and the engines have not come loose in a couple of decades of very limited use. That said, I included 1 inch dia x 1 inch long spacers under the forward engine mounting holes - back two attachment holes are now directly under the axle and I did not have adequate room for such a spacer. I did torque the SHCS instead of just "good and tight" the first time.

I like your coupler solution

[rickyb]

Beep beep beep, back it up a but.
I am an engineer too, but you are getting ahead of yourself.
Everything you said is probably true. Yes indeed.
But step back and see the big picture.
Single cylinder engine, lots of instantaneous torque pulsations,
I have first hand experience with more than a few Tecumseh 8 and 10
hp engines. They vibrate a lot. They break gas tank brackets regularly.
Ever use a snow blower with one? Holy crap they vibrate a lot.
You want to sound like an engineer, talk some about overhung load,
and you would be closer to the root.
But bolt stretch, not a factor here, or so small a contributor that it is
really a distractor. Not ideal, but bigger sins are causing his failure.
Don't give machinists an excuse to pick on engineers.

--D

Well Dooz, there is method to my madness. I was just working within the original constraints. Substantially improve on the 7 hour changeout without major mods. Fixing the fasteners would in my mind shift the failure mode. The next failure would be the aluminum casting but the fasteners wouldn’t fall apart. Thought that might be a big enough, simple enough bandaid that meets the original goal. Of course getting rid of the vibrating noodle and decoupling is the real cure. I think the OP has a nice final solution.

[Doozer]

Fair enough. --D

[Willy]

I like your execution of an alignment sleeve, good thinking.
Also the entire assembly of components just seems to follow a best practices approach.
Not always easy to tame the "paint shaker with a spark plug" that Tecumsehs are known for but as you say one gets a good feeling after an hour of run time as to how much better it now behaves.