How to connect two shafts 5

[Matt51]

I want to connect two shafts directly to each other. They are small diameter and have large axial loads, they push and pull, but no torque. I can thread them together, but the required torque preload to prevent separation is very large. I am trying to eliminate a ball joint, which has wear. I do not have room for a flange with bolts. Are there any other choices?

 

[dvd]

Do the shafts need to be round?

What is the magnitude of the forces?

 

[MintJulep]

To someone who designs ships a "small diameter" shaft may be 2 feet, and a "large axial load" may be several tons.

To someone who designs watches, a "small" shaft may be less than a millimeter, and a "large" load a few grams.

 

[byrdj]

slip on sleeve over the shafts, then drill and pin through sleeve and shafts.

If connection needs to allow for mis alignment, then a "H" clevis with the ends at right angles, again drilled and pinned (the pin could be a through bolt)

 

[eromlignod]

Mate the two rods male/female and use a taper pin.

Does it have to be disassemblable? If not, you could thread both rods, screw them into a coupling, then weld it.

Don
Kansas City

 

[rb1957]

can you move the interface between the two rods to a different position (where you'll have more room to work with), ie change the length of the two rods ?

otherwise i like byrdj's idea of a sleeve

 

[jistre]

Weld them?

 

[CoryPad]

If it doesn't need to be disassembled, then use friction welding.

If you thread them, why do they need high (any) preload? The idea with high preload in fastened joints is to prevent high stress amplitude in the fastener, but if your shafts already see the high amplitude, then just thread them and use low preload (enough to remove allowances/tolerances).

Regards,

Cory

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[Matt51]

Thanks everyone for the excellent suggestions. Yes the parts have to come apart. Two inches maximum od. 15000 lb push pull.

I guess I agree Cory, sounds logical. Sleeve with pins sounds good also.

 

[MadMango]

Cut both rod ends into Ds with 2" overlap, and pin the connection.

"Art without engineering is dreaming; Engineering without art is calculating."

Have you read faq731-376 to make the best use of these Forums?

 

[EngineerTex]

I agree completely with CoryPad. Preload isn't necessary except for the amount required to prevent it from loosening on its own. A single connection like this doesn't operate under the same principles as the preloading concept of a bolted connection.

Oilfield workstring tubing undergoes exactly the same situation that you're talking about. It gets both pushed and pulled repeatedly during a well servicing operation and a well-designed and properly-torqued joint doesn't ever come undone.

2 inch OD with 15000 lbs seems like you have plenty of room to play with. Using standard ASTM SA53 SCH 40 1-1/2" nominal OD black steel pipe with 35,000 psi yield has an OD of 1.900 and an ID of 1.610 and results in a stress of only 18ksi and a SF of 1.86 against yield. SCH 80 gives you 14ksi and 2.5 SF. A tapered thread joint can come very close to 100% efficiency. Of course, buckling calcs may show lower SF.

To use 1-1/2" nominal pipe, though, you would machine a tapered-thread plug to connect the two on the ID as opposed to how it's normally done, which is on the OD.

Use the removable Loctite.

-T



Engineering is not the science behind building. It is the science behind not building.

 

[rb1957]

since you have a tube (not a rod), you have several choices about sleeve design ... external, internal, or both.

for pins, if you can't fit standard bolts (and nuts), you could get a two piece pin, with one 1/2 having an internal thread.

 

[CoryPad]

I really think threaded ends is a more elegant solution. I think a sleeve with pins sounds horrible.

Cory

 

[jistre]

Hmm. If you're worried about losing preload, could a chemical locker used with a threaded connection provide the security you'd like to keep the assembly from coming apart until you want it to?

 

[dvd]

I hope this is an engineered shaft. 15,000 lbs will definitely buckle a shaft. What is the distance between shaft support blocks, and where does the joint occur relative to the support blocks?

 

[sreid]

Can you put grooves on the OD of the shafts and connect them with a grooved split collar?

 

[TomFin]

If you are concerned with buckling you can always brace the shaft at an intermediate location. The shaft then would have to buckle in the second mode, which will increase your critical load.

Failure is a prerequisite of successful design

 

[rb1957]

i think we can all (well, most) figure out an euler column load, and adjust I to suit.

one thing to think of is the connection between the sleeve and the tube (shaft) will need to transmit moment as effectively as the shaft; ie, if there is moment flexibility between the two shafts then they won't act as a single shaft (in compression), and two shafts (joined by a hinge) would have a reduced euler load (no?)