Shrink fit stress calculation on slotted shape

[bjason]

I have a 5x2 retangular part with a 3" x 0.5" radiused slot that I'm heat shrinking onto a boss 3.006 x 0.5
How can I calculate the stresses in the joint?

 

[MikeHalloran]

For a very rough start, do the calculation for a shrink fit of a 0.506 rod into a .500 hole.
In any practical material, the stresses will be very high, but all it really tells you is that there will be some distortion at the ends.

With the dimensions as stated, or as I misunderstand them, the only stress on the rectangular bars connecting the ends will be tensile, from stretching each bar by .006" or so.



Mike Halloran
Pembroke Pines, FL, USA

 

[bjason]

Mike
You are correct regarding the "bars" being only in tension.
Yes I did the calcs for a .506 shaft in a hole. Im thinking that with the slot design the stress at the ends will go down?
What I'm looking to know is how much interference fit can i have before material yeilds? the tool steel Ive selected can expand up to .014" over the 3" length with out getting into tempering range. The .006 was a starting point based on a 3" diameter
I want this joint to be strong as possible and still be able to release the joint with heat and replace the slotted part. ie the boss needs to be reusable. I know getting the parts apart wont be easy but thats the next step after this - we will setup a test jig for this.

 

[MikeHalloran]

Right, the pin in hole gives you a high estimate.
But the pin load is applied at the weakest plane of the 'crossbars' at the end, and they'e stressed in bending, so they'll be 'pinched' a bit.
If the slot were rectangular, the problem would be easier to analyze, but the assembly wouldn't be self-centering, which I assume is a requireement.
That's about where hand calcs start to make my head hurt.



Mike Halloran
Pembroke Pines, FL, USA

 

[rb1957]

but isn't the 6 thou interference on the 3" length (not the 1/2" width) ?

i'm not sure about the disassembly, unless the boss material has a lower coefficient of expansion ?

why not use screws to hold things together ? maybe local the two parts with an interference fit spigot ? certainly disassembly would be easily, and reassembly too i'd think.

 

[bjason]

Yes the .006 interference is on the 3" length
Joint will seperate with localized heat and appropriate force.
Threads create stress points and require more machining.
Please say more about the interference spigot idea - thats basically what I'm proposing?

 

[rb1957]

rather than having a slot in the green piece, i'd have it solid with two or three holes in it. there'd be matching holes in the supporting grey piece.

install freeze fit pins to join both pieces.

to disassemble, drill out (carefully !) the pins. possibly repair the holes with bushes (or use larger pins)

 

[MikeHalloran]

I think maybe he's trying to come up with something like the shrink fitted/ heat released tooling used in some high precision high speed CNC machines, where the ordinary cold strain in threaded parts is not considered reliable enough for retention at very high rpm.

Depending on the axial depth of the tooling, the proposed scheme may not be particularly resistant to external moments applied in certain planes. Perhaps the problem is as asymmetric as the solution, and my concern is ill-founded; insufficient information was provided, no doubt for commercial reasons.



Mike Halloran
Pembroke Pines, FL, USA

 

[rb1957]

pretty much that why i offered the idea of spigots

 

[bjason]

Mike you are right it is a tooling design. Weve played with the pin design but ran into machine tol issues getting them to line up.
All the loading on the joint is either from the top or left end of the green piece up to 130000lbs. Weve had some success in field trials and now I'm trying to refine quantify the design.

 

[dhengr]

Bjason:
Until you posted that sketch I couldn’t make heads or tails out of what you were talking about, for the same reasons rb1957 and MikeH raise. When you describe something in words only, you must realize that none of us can see what you are looking at or imagining. Then read your words aloud and imagine all the different arrangements that anyone could draw from those words. If more than your intended detail could possibly be drawn, your word description is inadequate. Provide a sketch, show loads, their directions, dimensions, etc. so others can start to visualize the full scope of your problem.

130kips pressing the green block down onto the gray base is probably no problem. 130kips bearing on a .5" wide x ~.5" high round nosed bearing surface on end of the .75" high boss is probably not so good, and the shrink fit will not overcome that deficiency. (130k)/(.5 x .5) = 520ksi in bearing is mighty high. In fact that .5" x 3" boss is probably not adequate in shear at the interface btwn. the gray and green blocks; (130k)/(.5 x 3) = 87ksi, a fairly high shear stress, probably applied more at one end than uniformly. The boss has to be large enough to resist these loads.

Why not have to boss on the underside of the green block, and just lightly press fit into a slot in the gray base, for proper locating, and load transfer. Then drill through the green block/boss and use a couple bolts to only hold the green block down. Lock these bolts with tab or wires, or some such, so they can’t loosen.