Bearing Stress on a Shackle to Shackle connection 2

[andy5powell]

Hi all,

I am after a little help. I have been asked from a client to assess the bearing stress on a shackle which has another shackle directly attached perpendicular to this, i.e. one cylinder crossing another, with a load less than the SWL of the shackle. Normally I would approach this using Hertzian stress or using Roark's however using these methods I obtain a stress 13 times higher than yield.

Obviously the shackle does not yield under this loading, so I am guessing either the methodology is incorrect or my limit of 2 x yield stress in bearing is incorrect. I am no shackle expert, so are there any other methods anyone can think of?

 

[boo1]

Consider the Hertzian contact stress. The pressure between two curved surfaces should be infinite, which would cause yielding of both surfaces. However, a small contact area is being created through elastic deformation in reality, limiting the stresses.

Bearing stress should be considered

https://en.wikipedia.org/wiki/Contact_mechanics

 

[MikeHalloran]

The shackle yields. The pin bends. Neither is designed, intended, or evolved for the stress situation as proposed.
Perhaps the customer could tolerate linking the shackles with a double padeye, sort of like a big roller chain link?




Mike Halloran
Pembroke Pines, FL, USA

 

[SwinnyGG]

Shackles are absolutely designed for and capable of being used in series. Shackle-to-shackle is no different with regard to local loading than chain-to-shackle.

The only rule that must be followed is that shackles in series should be linked bow-on-bow or bow-on-pin, but NEVER pin-on-pin.

 

[Lomarandil]

I don't have a good reference to back this up, but I've also used shackles in series as jgKRI mentions.

My understanding is that for Crosby shackles at least, even pin-on-pin is acceptable as long as the pins are actually bearing on each other (the shackles aren't being bound up at the ears).

 

[andy5powell]

Essentially what I have is a master link going around a pin, so as a cross section, it is a pin over pin. I have seen this before with testing etc where they have shackles and the like in series, just I've never managed to calculate the bearing stress on the shackle and prove it is fine.

 

[SwinnyGG]

If you can't flip the shackle so that your 'master link' goes through the bow and is thus self-centering, the next best alternative is to shim either side of the pin (with washers) so that the 'master link' can't apply an off-axis load to the shackle.

Also, in my general practice, I try to never exceed SWL/2. Obviously SWL can be approached safely, but in my industry I often deal with the 'that looks strong enough' mentality from millwrights, so I use that rule of thumb to try to spec in an additional layer of safety against stupidity.

In other words, consider going up a shackle size or two, if you can.

 

[dhengr]

Andy5powell:
The localized bearing stress in a situation like yours is well confined by the surrounding, non-yielding material; so, it is typically allowed (tolerated) that it be much higher than our normal allowable bearing stresses would be. The plastic deformation occurs, the bearing area grows until there is localized compatibility and equilibrium, and it really can’t/won’t go any further because of the larger volume of material containing it. It works. As mentioned above, the initial contact stress is essentially infinite (as your calcs. tend to indicate) and small mating bearing areas are formed and grow in size which quickly brings the bearing stress down to a reasonable level. This is indeed a Hertz contact/bearing stress problem, and you should look at it from that standpoint. The other comments above are to suggest that you want the shackles to act as they were intended to act, and in the line they were intended to act on. JgKRI’s 30JUN16, 16:04 advice is particularly good. I would add that pin-on-pin contact puts the pins in too much bending, and they are primarily intended to work in shear (double shear). Then, there is also the chance of some shock loading if one pin slides on the other pin as loading occurs. The dia. and size of the shackle body are designed to tolerate a bit of this type abuse (deformation and wear), while the pins are not.

Edit: You could cut a piece of heavy walled pipe to fit over the pin, pipe i.d. slightly larger than pin dia., with its length to fit within the two sides of the shackle. This protects the pin from damage, distributes the master link point load on the pin and with the pipe o.d. being larger improves the dia. variables btwn. the pin/pipe and the master link in the Hertz stress calcs.

 

[swearingen]

As another vote to the "it'll work" group, as an intern in college, I was the company witness for offshore shackle tests. They took all shackles used for lifts (from 20 ton to 600 ton SWL) and put them in a pull-tester bow to bow, then ran them up to 2xSWL.


-5^2 = -25 ;-)

http://www.eng-tips.com/supportus.cfm

 

[BUGGAR]

A news agency once got ahold of some fem printouts for a bridge showing big red patches and overstresses at some major pin holes. A semi truck driver couldn't back up better than they did.

 

[andy5powell]

Thanks all. Unfortunately in my industry I also get the 'its been fine for 20 years why change it' mentality, even though the problem has never been proven. I guess its more wide spread than I thought

 

[spieng89]

Link

The above link from a manufacturer explains the limitations of linking shackles. It is allowed but under certain conditions. A call to the manufacturer of the shackles you are using may be helpful as well as they may have literature on testing linked shackles.

 

[Lomarandil]

Good find raspivey!