For the design of Spreader and Lifting Beams, The CLAME code seems to give a Safety factor of 2.6 on failure in buckling 1.3 x proof Load of 2=2.6). (Section 4.2.3). This seems way too low, when normally I would expect around 5 on SWL, especially for Offshore type lifting where we have Dynamic Acceleration factor from sea state and lifting factor from rate of hoisting.
Opinions and comments would be appreciated.
I've never used Lloyd's, but a SF of 1.3x2 seems low to me as well. A bare minimum of 1.92 (23/12 from ASD) for static buckling has been my course of thought in past offshore spreader designs and then putting a dynamic lifting factor on top of that of 2.0. So the 2.0 on top of the 1.92 would give me a final factor of 3.84 against yield vs. static loading.
I don't agree with the 5:1 blanket SF for offshore except in wire rope slings, shackles and live lines on the crane itself.
Additionally, even if Lloyd's says 2.6, I don't think they intend it as a maximum.
Engineering is not the science behind building. It is the science behind not building.
A factor of 5 is arrived from the following
Buckling is an Ultimate failure, so should be compared with Ultimate strength requirements. What I understand(and I may be wrong) is that Lloyds calculates 2x1.3=2.6 Safety Factor on Ultimate where as API Spec.2C looks at the following (1.33x 1.5)/0.66 =3.0 with respect to yield.
If we use a 350MPa steel with 480MPa Ultimate, this gives,
3.0 x(480/350)= 4.11 SF as a minimum on failure, using API2C, (I use 5).
Buckling is an ultimate failure, but it occurs when stress crosses yield. You cannot consider ultimate strength when dealing with buckling. Not for ASD and I doubt that it's true for LRFD, either. Neither Johnson or Euler (or the equations combining bending with buckling) consider ultimate strength. These equations are based on elasticity and yield or elasticity alone.
After you hit yield, you're no longer in the elastic region of the slope and from that point, things go downhill in a hurry.
And:
Think of this: In the theoretical world, consider a system where a mass is supported and there is an elastic line with no slack whatsoever attached to the top, which carries none of the weight of the mass. Then, if you instantaneously remove the support, the elastic line will see a total load (for a brief instant) of exactly 2.0 times the weight of the mass. This is similar to what a crane (or lifting sling) sees when dynamically loaded. Your dynamic factor should be 2.0 and your static safety factor should be 1.5, 1.67 or 1.92 (tension, bending and buckling), resulting in an overall SF of 3.0, 3.33 or 3.84 against yield, just to name a few general cases.
I don't have API 2C in front of me, but remember that for the crane itself, a 1.33 dynamic factor is applied to your numbers to consider dynamic loading, though oftentimes, it is not the same type of sudden loading that a smallish skid would see. For a 100-ton crane lifting a 100-ton load, consider the elasticity of the entire structure, including the platform in addition to the speed at which the supply vessel is heaving. A 100-ton load is less able to apply a sudden impact loading to the crane due to the physics of the overall system. A 100-ton crane, however, can very easily apply an extremely rapid loading to a 5-ton skid being lifted from a heaving vessel, where the skid's lifting slings will see much more than 5 tons. So a 3.0 - 4.0 SF against yield for the skid structure and solid parts is reasonable, although the 5.0 SF against breaking for the wire ropes should also be used. The 5.0 SF against breaking strength of the wire rope is used because it's extremely difficult to determine a yield strength of wire rope.
Furthermore, we're not putting up buildings. For machinery, yield *IS* failure. The moment that parts have yielded, they no longer fit/work together as originally intended. Throwing a safety factor of 5.0 against the wrong failure doesn't solve the problem.
Engineering is not the science behind building. It is the science behind not building.
Just a couple of comments about Buckling.
I am talking about structures, lifting beams, (not machinery).
Local buckling (crippling), is section area dependent and will/ may not lead to collapse, or failure of equipment.
Column Buckling is L/ry dependent and will lead to collapes, failure. It will always happen at below 0.2% offset yield stress.
For High value L/ry> 50, Structures, where we use Hollow sections, L/ry is very often most critical. For I-beams it can be anything (local crippling or column buckling).
For Ultimate State Design of equipment (LRFD) (such as offshore rigs) there are both Servicability Limit State, where equipment has to work after load is applied, and Ultimate Limit State, where plastic bending is allowed (no snap).
I think that if you use a SF of 3 against yield, and the structure can still carry load i.e doesn't break, snap, release the load completely, but bends plastically, then it adds up to 3x480MPa/350MPa=4.1 against failure, snap, release load. Buckling is the mode of failure where there is snap and release of load, therefore the factor should be at least 4.1 (say 5).
I have no problem with the platform itself and all its beams and underpinnings being designed to LRFD to include minor plastic deformation.
However, if you're talking about a spreader bar for a lifting appliance, you had better not let it permanently deform. And I still see no place for the ultimate strength to be considered in the analysis of said beam.
Engineering is not the science behind building. It is the science behind not building.
