Tricky tailing lug design -NO WRC 107

[abehong]

I am currently designing a tailing lug for a pressure vessel repairing project. The client wants to replace the top portion of a vertical tower ( partial of shell and top head). The shell thickness is ½” but with 4” refractory internally installed . The removed portion will be lifted up using existing lifting lug installed on the head and lower to ground level. It is request to design and install a tailing lug at bottom of removed component , so the rigging company could use it to lay the removed component down ( from vertical to horizontal position) . The design load on tailing lug is fairly heavy ( about 138,000 lbs unfactored ) because the tailing point is really closed to center gravity of removed component. Due to this heavy load, I am struggling on make local stress passed using WRC 107 method (it actually seems to be impossible to make it passed). The WRC-107 method may not be the best approach to this analysis . Tom gave a detailed explanation as per discussion before in this forum.
Why not? Two reasons:
(1) WRC-107 determines stresses in the "free field" of a shell; the effect of any local shell stiffening is not considered. The tailing lug may be located very near to the skirt base ring, or even attached directly to the base ring, in which case the assumptions of the bulletin are not valid and the analysis may be overly conservative.
(2) The bulletin is limited to rectangular attachments that do not exceed a rectangular aspect ratio (eg: length to thickness ratio) or "footprint" on the shell of 4:1. ie: The length of the tailing lug along the skirt cannot be more than 4 times its thickness; most real tailing lugs will be longer than 4 times the thickness. This doesn't mean that the actual lug can't be longer than 4 times its thickness, but the analysis by WRC-107 will be limited to this length. Thus the analysis will likely be very conservative or even overly conservative. Not a problem if it doesn't force you to modify the construction. But if it shows as overstressed, then it may unnecessarily force you to some more expensive construction.

I like to get some comments for two clarification. First, Is there other simplified methods to check local stress in addition to FEA method for my case , eg ,some design criteria in Roark’s stress and strain ? Second since customer will abandon this cut off portion of shell, I think that it should be reasonable to adjust allowable stress to be same as yield strength of shell material . In another word, material yield is not big concern for this design as long as it not crashed . The safety margin of the ultimate strength over the yield strength is considered sufficient to provide the required safety factor. I am not sure if my thought is accurate or not.

Thank in advance for any feedback

 

[JStephen]

Can you put two lugs 90 degrees from your intended location so they are loaded mainly in shear?

 

[TGS4]

If you are dealing with lugs near or on the base ring, I could only recommend and FEA - the interactions between everything would be very challenging to add together.

Regarding the "allowable stress basis", I could not recommend any value in excess of the Code design values. Why? Because the localized membrane stresses will already be limited (in the evaluation of the FE results) to essentially yield. Plus, the membrane+bending stress will be limited to twice yield. You won't really gain anything by bumping up your allowable basis, except invite further scrutiny. There may even be some who may argue that ASME B30.20 may apply, or even your local OH&S rules, which are far more restrictive. It's true that you will scrap the vessel after the lift, but you still need to maintain an adequate margin during the lift. The last thing that you want is for something to fail under-the-hook; and you certainly don't want to spend the next four years of your life in a lift failure investigation.

 

[MJCronin]

Pictures and sketches would help here....

I suggest that you consider two tailing trunnions with repads located at 90 degrees apart..... they will be able to take the load.

You will need a spreader beam at least the diameter of the vessel.

Also consider the use of adjustable chainfalls during the rigging process.

 

[MikeG7]

Ditto what MJCronin says. Trunnions are the way to go. Will distribute the stresses better with the pads.