Two shafts assembled with glue

[Tmoore]

Does anyone know how to calculate the following. I am going to be assembling two shafts together using a shrink fit, coating the interior of the two shafts with loctite, then welding the mating point of the two shafts with a fillet weld. The outer shaft will be attached at the bottom to a gear box, then the inner shaft will attach at the top to a coupling, as the top of the inner shaft has a coupling as well. I need to figure out how to determine whether the inner shaft will slip when the gear box shaft starts turning the outer shaft. I know I have not given any values, but I was wondering if anyone has ever had to perform an analysis similar to this. Any help is welcome.
Cheers,
Tom

 

[FrenchCAD]

wow, do you really need such a complicated design? I have no idea if you can estimate this sort of things.

Cyril Guichard
Mechanical Engineer

 

[GregLocock]

This sounds like desperation. Been there done that, at least you haven't got a Woodruff key in there as well.

OK, quick answer is no. Long answer is trust in Loctite. 600 grade will give you a welding type joint. If you mix a weld and a loctite joint then you will exploit the advantages of neither system.





Cheers

Greg Locock

 

[GregLocock]

Ack, the machine ate my post.

Combining a weld with a loctite joint is not really a good idea. If either is good enough then the combination will be no better, and if either are inadequate then you will be lucky if the other can take up the slack.

Usng Loctite 600 grade you should be able to generate a joint that will hold together. Loctite will guide you through the calcs.



Cheers

Greg Locock

 

[hacksaw]

use of loctite and welding seems to be an incongruous solution. of course you can always try it in a pinch, but usually you need to identify the torque loads (starting / running / stopping / overload , etc.), axial thrust you are designing for and the degree of mis-alignment, etc.

rarely do you connect a gear box to a load without some sort of a coupling as opposed to loctite and welding.

 

[Tmoore]

Thanks guys.
I was feeling very discouraged that I could not figure this problem out, but it is encouraging to find that it is as difficult as I first thought. I have very little information that helps me. I do have the torque that the motor supplies, it is 40 N-m. Then, because of the gear box it becomes 200 N-m. I also know the outside diameter of my inner shaft, that is fixed as I am buying it from a supplier who cannot modify the part. Therefore I can select the inner diameter of the outer shaft, and therefore decide how tight the press/shrink fit will be. I have a method available to me from my solid mechanics notes that helps me calculate the stresses caused by a shrink fit, but they are very general and do not go into enough detail for me to confidently solve this problem. I have looked into using the loctite and was going to use the 635 compound. It can accomodate up to 3000 psi of shear but I am having a bit of time wrapping my head around how to relate my torque to this shear stress.

My best thought was take the torque, convert it to a force at the mating surface, and say this is my friction force, then using the coefficient of friction of the two surfaces, find my normal force. Then this normal force would be the force applied by the pressure caused by the shrink fit, and this pressure would be my axial pressure. Then I would use this pressure value and my solids notes to solve for the shear force.

I was also wondering how many of you think that I should just use one shaft. I cannot think of any advantages to using this press fit shaft system. Anyone have any advice?
Thanks for all your responses...
Tom

 

[prex]

If I understand correctly your assembly, the shear stress you are seeking should be
&[ignore]tau[/ignore];=2M/&[ignore]pi[/ignore];D[sup]2[/sup]L
where M,D,L are the torque, the diameter and the length of the mating surfaces.
Concerning your last question, I see that everyone here suggests you not to use such a messing joint (press fit + glue + weld??): however you are the only one that can propose alternatives meeting your project constraints.

prex

http://www.xcalcs.com

Online tools for structural design

 

[Tmoore]

Hi prex,
I am actually not in charge of the design, only in charge of the calculations. My boss feels that using all three of these methods of attaching the shafts is best, I am just trying to make sure it will work. I was wondering, in the equation you sent me, what does n represent?
Thanks
Tom
P.S. Personally, I want to make it all one shaft...as there is no advantage I can think of to using two shafts.

 

[evelrod]

Ahhhhhh! You have my sympathy Tom. The old "pointy haired boss"!
Using LockTite in combination with heat is not a good idea as expressed by others. 6xx works well in bearing installations but I have some reservations about a 'locktited' shrink fit connection.

Rod

 

[Tmoore]

Does anyone have any comments with regards to using one shaft instead of two? To be totally honest, I don't want to put my trust in loctite. These two parts, should they fail for any reason (ie. the loctite fail), would take several hours to replace. I do not see how the single shaft could fail.
Tom

 

[evelrod]

From your description, Tom, I really don't understand the need for a two part shaft if a one piece will work better. Then again I am not the "pointy haired dude" either. I do have some reservations about connecting a gearbox output shaft to anything without some sort of compliance joint.

"I do not see how the single shaft could fail."
Your the engineer, I guess it's up to you to make your case.

Rod

 

[prex]

'n' is [tt]&[ignore]pi[/ignore];[/tt](pi)!

prex

http://www.xcalcs.com

Online tools for structural design

 

[Pekelder]

Welding and loctite will almost surely leave the loctite hamred, and thus you will end up with a welded coupling. So, when welding, skip the loctite.
I've used loctite for connecting gears to shafts, works fine if you do the calculations right, but bear in mind that in case of overloading/shockloading, the glue might fail, and the system will slip. If you're sure about the calculations and conditions, loctite (or other brand) will perform well.
You could use a clamping-bush to connect two shafts, even of different diameters.
Welding a shaft might give couse to fategue if the welding is poor. The shaft cannot be dissasembled (loctite can be heated to get it loose, but in most cases, with the surrounding machinery this is not possible). If it cannot be disassembeld, why use two shafts in the first place (you be the judge of that..).

Regards,

Pekelder

 

[Viperin]

Hi
have you already think of instead of gluing the shaft you could join them by a heat shrinking joint(heating one shaft so it expands an the insert the other shaft in it.


best regards

 

[viscomike]

My company experimented with Loctite.

We found under the reversing loads, the kind we have all the time, it failed rapidly.

This was 10+ years ago so maybe it is better now.

When we have "cannot afford to fail" applications we use a key and heavy shrink fit.

Our key has a slight interference with the keyway so that it must be pounded into the slot -- no jogging if the shrink fit fails.

Machinery Handbook gives recommended shrink fits.

 

[IRstuff]

Another thing to consider is the possibly of compromising the weld with the Loctite, e.g., if it flows back out or outgasses, it migh interfere with the making of a good weld joint.

TTFN

 

[automatic2]

How do you break a Loktite joint, by heating it. Forget the Loctite. Custom drive tubes with or without welded couplings are fabricated by weld. The shrink fit is not for transmitting torque, but for ensuring alignment. Make sure your tubing has appropriate wall thickness to transmit load, and house the weld.

 

[PJGD]

For a somewhat similar critical joint situation, we use Trib-gel:-

http://www.tribtech.com/prod01.htm

It may be beneficial in your case too.

PJGD

 

[Executnr]

Tom
'Mark's Hanbook for Mechanical Engineers' has a pretty good set of equations for the amount of torque an interferance fit can take.

I'll try to get the particulars together if you want.

Let me know
Wayne