Lifting Plate Design Assumptions 1

[CARunderscore]

Hello, all. I don't post much, but I've been a lurker for a couple of years. I've been tasked with "verifying" that someone else's lifting plate (or lug, if you prefer) design will work for another application. However, upon looking at the original calculations for the plate, I'm not totally sure that it is really adequate for the original intended use. Also, the design just looks... odd... to me (see attachment), and I'd like to hear some other opinions before trying to size it up for a potentially smaller load.

Right now, I am just concerned about the vertical load case, although the plate will need to support load through a 180-degree range of motion. The original static equivalent design load was over 200 tons and assumed to be equally shared between two identical plates with a spreader beam such that there would be no appreciable no out-of-plane bending. Checks were/are to be performed based on ASME BTH-1-2014 Service Class 0 Design Category A, with an additional reliability factor of 1.2 for an actual Design Factor of 2.4. The shackle pin is about 4" in diameter. There are some interesting checks I will have to perform on the base member, but it is the area around the pad plates that I am focused on.

I have a couple specific questions, but I'm open to any comments about other aspects of the design. Sorry if I'm a little vague about numbers. I am at home right now, and I'm working completely from memory:

[ol 1]
[li]Is it valid to consider the section around the pin hole as a single 3" thick plate with a 6" outer radius, conservatively neglecting the outermost inch of the center plate radius? I was under the impression that these pads are often designed only for bearing stress, and not relied upon to contribute to the pinhole strength. If they are considered for strength, then the welds need to be adequate to transfer load back into the main plate. I'd think that the effective weld length should be considered to be half of the pad circumference at most, because above the pin hole the pad is supposed to be reinforcing the main plate. I don't see how the upper half of the weld could be a valid path for net stress between the pad and the main plate.

I found an old online calculator that seems to use the full circumference (

http://maximumreach.com/PlateLug.asp),

but it also uses an impact factor of 1.80, which I believe is equivalent to BTH-1 Category B. The higher design factor makes me less picky about load path assumptions. Also, I do not think that calculator was made with pad plates this thick in mind. The plates make up 2/3 of the total connection thickness, and they are both 3 times as thick as the adjacent fillet weld leg.

As a side note, the original design calculations stated that each of the two 5/16-inch fillet welds with 38-inch circumferences had an allowable shear strength of 1,750,000 lbs.[/li]
[li]What is the net section I should consider for tensile yielding immediately below the pads? The change in thickness is rather dramatic. The senior engineer I am working under directed me to use a Whitmore section-like approach, which seems logical, but where should I start my included angle for a plate with a single large hole? The center of the pin hole or the top of the pin hole seem conservative to me. The top or side edges of the pad plates seem reasonable, but give fairly different answers. Or should I just be checking a failure plane that follows the bottom edge of the pad plates? Something like block shear?

I'd think that my assumed failure plane should correlate with the effective weld length from my first question. The original design calculations assumed a tensile area of 24 sq. in.[/li]
[/ol]

I am inclined to size the pad welds up. I also believe that the original designer miscalculated the strength of the plate, but I think that the thicknesses may still be adequate.

Sorry about any incoherence. It's rather late for me.

 

[JStephen]

I was thinking that the BTH design guide actually indicated a couple of different ways the reinforcing pads could be handled- and one way was calculating the strength of each plate separately and then adding them together.

In the examples I've looked at, the loading on the welds turned out to be relatively low, so that wasn't a big issue.

Will the design work just assuming the diameter of the circle as the net width below?

Edit: See the Commentary following 3-3.3.1 (in the 2008 version, at least, may be different in the newer version).

 

[CARunderscore]

See the Commentary following 3-3.3.1 (in the 2008 version, at least, may be different in the newer version)

I have consulted the commentary. In both the 2014 and 2017 editions, it reads as follows:

"Pin-connected plates may be designed with doubler plates to reinforce the pinhole region. There are two methods commonly used in practice to determine the strength contribution of the doubler plates. In one method, the strength of each plate is computed and the values summed to arrive at the total strength of the detail. In the second method, the load is assumed to be shared among the individual plates in proportion to their thicknesses (i.e., uniform bearing between the pin and the plates is assumed). The method to be used for design of any particular connection shall be determined by a qualified person based on a rational evaluation of the detail."

Both methods give the same answer for the case of equal plates. However, based on a "rational evaluation of the detail," I do not think that the upper half of the weld circle does much except provide a clean joint and potentially prevent dishing in thin plates (I hope I am using the word "dishing" correctly). I think that the welds need to be sized based on a length less than the full perimeter to assume a 100% strength contribution from the pad plates.

In the examples I've looked at, the loading on the welds turned out to be relatively low, so that wasn't a big issue.

Will the design work just assuming the diameter of the circle as the net width below?

Basically, my problem. This plate has odd proportions and lower safety factors compared to examples with pad plates that I've seen (and a higher base metal strength relative to the weld filler metal, I think), so I'm skeptical of just replicating a simplified approach without scrutiny.

If I take the area around the bottom half of the pad perimeter (pi*(12"/2)*1" = 18.8 in^2) and multiply it by 0.8 (interpolating between 1.0 and 0.6 for a mixture of tension and shear in plate... I think 0.85 is the true average factor based on calculus), then yes, it works, just with both a higher calculated stress and a lower allowable stress than calculated in the original design. I like this approach. Is is rational, or am I overthinking this? The difference might become important if someone tries to apply this to a heavier application at some point.

EDIT: Also, the thickness of pads relative to the main plate suggests to me that the main plate should just be thicker.

 

[SAIL3]

the OP is certainly correct in the questions proposed....going from a 3" thickness to a 1" looks odd at first glance...
the easy part......hertzian contact for 3" thickness with 6" outer radius...probably ok...
the difficult part......as mentioned, what cross section to check for tensile stress??...I would check a few cross-sections to try and determine the controlling case ....a few listed below...
CASE 1/....In this case I would try and envision a probable overall tensile failure mode....so what is the cross-section where I know with certainty that the total load has been transferred to the single 1" pl.....this gives a cross-section at the bottom of the doubler pls..either straight accross or following a portion of the circumference of the doubler pl and then straight to the edge....
The big unknown is the location where the majority of the load in the doubler pls could be transferred to the 1"pl other than in CASE 1 above or does it really matter....
CASE 2/....cross-section thru center of hole.....
assumption A..all (3) 1" pls still carry equal tensile loads...check capacity of lower 180 deg of doubler pls to transfer their load into the 1" pl...this would conservatively bound the problem and bypass the issue of how much load the top 180deg doubler pl weld imparts to the 1" pl, basically, assuming no weld for the top 180deg.....check, if ok, then I am not concerned about how much load has bled into the 1" pl at this section as it has an alternative path....
if not ok....I am in deep doodoo and would have to investigate/assume how much load is being transferred by the top 180deg of 5/16" weld to the 1" pl....I would initially prefer
to check it as not having any weld in the upper 180deg as it gives a lower bounds to the problem
CASE 3/...cross-section @ bottom of hole assuming 100% of load in 1" pl...or 75% etc...
.....sorry for the long-winded ramble

 

[SAIL3]

on further thought to get rid of the clutter...

A/..Bearing /hertzian stress....(3) 1"pls, 6"R....check
B/..Tensile stress....ignore doubler pls...check only 1" main pl...if ok, you are done..

I have not run any numbers but suspect the doubler pls may have been added to take care of just the bearin/hertzian stress.....I have also added them in the past to force the erector to use the right size shackle.....

 

[CARunderscore]

on further thought to get rid of the clutter...

A/..Bearing /hertzian stress....(3) 1"pls, 6"R....check
B/..Tensile stress....ignore doubler pls...check only 1" main pl...if ok, you are done..

I have not run any numbers but suspect the doubler pls may have been added to take care of just the bearin/hertzian stress.....I have also added them in the past to force the erector to use the right size shackle...

I have heard of the shackle-size-forcing trick before, and maybe that's a contributing factor, but the original was definitely designed utilizing the pad plates for actual end-area strength. The plate is at nearly 200% of allowable if you ignore them entirely. Which I would've been inclined to do, except that saying the plate needs to be twice as thick would raise eyebrows. I've seen less weird-looking lifting plates that also relied on the pads for strength through the pin hole, just not as heavily.

I am trying to figure out now what the actual constraints are on thickness, but now I am more worried about the attached structure and am beginning to think the whole design concept is ill-advised. There are four existing plates, and somebody wants to replace them with two stronger plates in different locations. Except that the existing steel is only 36 ksi, and the structure may actually be weaker at the proposed "better" lug locations.

 

[12345abc6ttyui67]

This looks like a fairly standard padeye to me, although the 'sloping' sides of the main plate are a bit odd.

I remember seeing a clause in one of the more obscure lifting codes (perhaps Noble Denton) which was along the lines of "cheek plates should not be more than 3/4 the thickness of the main plate in order to retain it's primacy". Usually the cheek plates are included in order to centralise a large shackle jaw on what would otherwise be a thin main plate and reduce the possibility of twisting / out of plane loads.

Don't quote me on the wording or the 3/4 limit, but perhaps there is a similar statement in the relevant code you are using?

Things I usually check:
-> Shear of the pin through the cheek plates + main plate
-> Bearing stress (Hertzian stress) of the pin against the cheek plates + main plate
-> Weld check of the cheek plate to main plate (I usually check for all load on one cheek plate which is conservative and rarely leads to a problem)
-> Tension over the 'narrowest' part of the main plate (e.g. a horizontal section just below the cheek plates)
-> Tension over a section of the cheek plates and main plate through the centre of the pin hole
-> Shear pull out of the cheek plates and part of the main plate from the rest main plate (e.g. a section round the lower half of the cheek plates, then extending vertically up the the radius of the main plate. Technically some of this failure is tension, but I just limit to shear which is more conservative).
-> Minor axis bending at the base of the main plate for 5% of P applied at the top of the pin hole

If P can also act an angle I also add the following:
-> Shear over the 'narrowest' part of the main plate (e.g. a horizontal section just below the cheek plates)
-> Shear over a section of the cheek plates and main plate through the centre of the pin hole
-> Major axis bending at the base of the main plate (for the horizontal component of P)

If P can also act at an out of plane angle (e.g. into / out of the page relative to your elevation sketch)
-> Minor axis bending due to lateral component of P

There are also usually limits on the hole size and cheek plate + main plate width in order to suit the shackle pin and jaw sizes, I'm used to hole diameter should be pin diameter + min(3% of pin diameter, 2mm), and cheek plates + main plate should be more than 75% of the jaw width.

I'll try and add a sketch on Monday as I have one on my work PC.

 

[CARunderscore]

-> Minor axis bending at the base of the main plate for 5% of P applied at the top of the pin hole

I had been thinking about performing a check similar to this, but there's no way it would pass at 5% of P. BTH-1 says that it does not require notional lateral loads, but they might be a good idea on a lug this thin. Not sure if I'd use 5%, though.

Result ended up being to just keep the existing lugs after all. Thanks for the input!