2 Point Sling Lift

[engpes]

I am trying to correctly analyze via FEA a 2 point sling lift in accordance with DNV 2.7-1. I have attached a diagram showing the boundary conditions that I have applied. 1 pinned connection at the single point lift, and 3 lateral translation restraints at the 3 frame corners shown.

This is the only way I can restrain my model to prevent instabilities and allow the program to solve. My question is this:

1.) Am I over restraining the model by preventing translation in the Z direction on the far left corner of the frame?

This seems to be the only way I can adequately restrain the model and my reaction at the far left corner seem to be small. I also tried pinned connections at all 4 corners and applied a single point load upward at the lift point, but this seemed to give results that were too favorable because the bottom of the frame was not able to translate.

Has anyone dealt with this in the past? This is just a load test that would be performed 1 time, so as long as the loading stays below the yield strength I am ok.

Thanks in advance for your help!

 

[engpes]

This was the 4 pinned connection I mentioned above. Thanks.

 

[msquared48]

I don't know what your loads, member stiffnesses or deflections are, but I would tend to install a temporary diagonal compressive spreader bar between your lift points to control inward deflection and a possible distortion of the frame of the structure.

Mike McCann
MMC Engineering

http://mmcengineering.tripod.com

 

[engpes]

Fair enough, but would you know of an accuract way to model / restrain this in an FEA?

 

[msquared48]

I would try to restrain it in X at two diagonal corners and in Y at the remaining two corners, then put in very weak vertical rod members down to four artificial supports that are restrained in all three directions.

If this does not work, you might try reposting in the RISA 3D forum. Josh could probably give you a better answer than me.

Mike McCann
MMC Engineering

http://mmcengineering.tripod.com

 

[Ussuri]

Are you trying to do a racking check on the container for lifting?

Lifting a container across the diagonals with a 2 leg sling is not a feasible lift. Its unstable.

Normally if you are trying to consider the effects of sling length and fabrication tolerance you would still analyse your lift with 4 slings. However, you put on two pairs of diagonals, each pair with a different apex node. You then fix one node and allow the other to translate vertically.

Now you need to do an assessment of load split between your diagonal pairs. You are using 0%/100% which I think is very conservative considering you cant lift it that way. 25%/75% is commonly used.

 

[rb1957]

it does look like a very odd way to lift, but that's your design.

the way i see the 1st pic, i think it's over constrained.
it looks like you have X,Y, and Z constrained at the lift point, yes?
then it looks like you have two translational constraints at two nodes, and another one at a 3rd node ... that'd make 5 translational constrants + the 3 at the pin = 8 > 6
i think you can delete the contraints at one node with your double constraints.

that said, i think your model will suffer from instability, particularly if you model the slings with rods. the problem, real and mathematical, is that the lift point is not well connected to the rest of the model.

further comment, what do you do with the loads generated at the model constraints ? hint, you can't ignore them.

 

[engpes]

It's lifted with 4 slings. This is for a DNV diagonal sling pull test. Similar to racking test. Thanks.

 

[rb1957]

do you mean that normally it's lifted with 4 slings, and you're modelling a test case with only two ?

my comments still apply ...
i think it's over contrained (if i understand your constraints),
i think the FEA will be unhappy with only two rods supporting it (if that's how you've modelled the slings).

isn't this something solvable by hand ? you know where the CG of the load is. assuming the CG of off-center, and assuming your constraints on the corners represent tethers, then the lift point reacts the weight, and the off-set (unbalance) moments are reacted by couples between the lift pt and the tether; yes?

 

[RockEngineer]

It's normal to design a 4 point lift to only use 2 legs unless you have a specific method of equalizing the load. Sometimes I will model a third leg going to one corner and then show there is no load in that leg from the results. For rotation about Y I would restrain one corner in x and one in z. I wouldn't restrain any of the corners in the y direction so all the vertical load ends up at the pinned support. You may have to fix your cables against rotation about their longitudinal axis. I usually use STAAD so it may be a little different in RISA. If you don't use the 3rd leg you may put a weak spring at one of the diagonal corners not supported to provide that stability.