Lifting Load Case 4

[phamENG]

If designing a temporary structure, or a portion of a structure that is going to be lifted during construction, how do you approach capacity checks? I'm specifically speaking about steel here, but insights into other materials could be valuable for posterity/future reference.

I know ASME has their below the hook lifting device manual, but unless I'm actually designing a BTH device or a fixed rigging point I don't like to use those factors of safety - they really punish you when you're designing something that will be moved once, maybe as many as 5 or 6 times before it's either bolted in place or thrown in the scrap pile. I don't need to design for 20,000 load cycles or even a 10 year life cycle.

Do you just use AISC ASD/LRFD factors with static loading and call it good? Good crane operators should keep accelerations close to 1g, so I'm not worried about dynamics. This feels like the most reasonable answer - what does everyone else think? (This would apply to any connections and members in prefabricated assemblies loaded in a manner they may not otherwise experience in service, using standard AISC equations for pins in padeyes fixed to the members, etc.)

 

[phamENG]

steveh49 - thanks for posting that. Excellent insight into the precast industry, which deals with this sort of thing all the time.

Since I'm ignorant of Australian codes, would you add an additional resistance factor to the 0.41*sqrt(f'c)? As is it's only slightly less than ACI's limit on flexure for plain concrete elements with resistance factor applied (equivalent would be about 0.51*sqrt(f'c).

As for load factors, I can ignore the suction factors working with steel and look at the dynamic and service life factors. I'd be curious to see how those were developed and the speeds they're based on as mentioned in the footnote, but the numbers themselves are certainly valuable.

 

[EngineeringEric]

PahmENG, If you go back into practice I hope we become colleagues and not competitors! (seriously if you come back to consulting, please let me know!!)

Mid-Atlantic. Virginia. Hampton Roads

 

[phamENG]

Thanks, Eric. I figured as much. I was on the southside before an industrial client brought me in house. I almost live in NC, so the drive all the way up to where you are may be a bit much, but I'll keep it mind should when/if the time comes.

 

[EngineeringEric]

We have a good deal of projects and clients (including industrial ones...) that way, but probably a bit to far north for daily commuting... But plenty of lovely houses for sale this way (my last sales pitch in this forum!)

 

[Settingsun]

The stress limit seems adventurous if cracking means failure of the hoisted load. It's written as a serviceability limit so perhaps there's an assumption that there's reserve strength from reinforcement. I don't see any capacity factor in AS3850. Plain concrete footings in our design code are limited to 0.6 (=capacity factor) * 0.6 * sqrt(f'c) = 0.36*sqrt(f'c).

I couldn't find limits on speed with a suspended load in AS2550.1. AS1418.5 says the maximum travelling speed should be given on a crane's load chart.

 

[phamENG]

That makes sense - limit the flexural tension to that level and you won't cause damage to the load. Thanks for clarifying the code references.

 

[Craig_H]

Pham, sorry to jump in so darn late. I've been preoccupied lately.

I used to work in the heavy lift and transport field. I breezed the above comments, and thought I would offer some input from the perspective of those lifting what you're designing. In my experience, the lifted load gets broken down a wee bit into "lifting components" and the "load".

The lift lugs that the rigging will hook into should be designed as lifting components. ASME BTH-1 is a great resource for designing these as simple pinned connections. I did see quite a few lift lug designs over time, and noticed that quite a few made the same common errors. Keep in mind that shackles are hot formed in a VERY loose tolerance environment. Stated dimensions have some pretty good variations, so I always made sure to give good clearances where possible. The common errors that I saw were:
- Insufficient room in the "bow" of the shackle. The lift lug plate would often protrude too far into the bow, leaving insufficient room for the shackle AND a sling to fit with adequate clearance. BTH allows you to calculate pin holes quite close to the loaded edge; use that ability.
- Shackle pin/hole fitment too tight. I once saw this designed as a mechanical fit, including the fitment category and everything. BTH allows for pretty good hole clearance, and I would utilize that where possible. The shackle pins are often not perfectly straight, so they can start to jam in tight holes.
- Insufficient room for the shackle eyes where the pin goes through. There was either insufficient room in between the eyes, or insufficient room for the eyes themselves.

In short, make it easy to hook 'er up.

IMHO, the remainder of the structure below the lifting points should be designed using standard LRFD principles, but considering rather flexible supports. Rigging is not a rigid support system like the structure's final foundation. It is often a system of slings and bars that can deflect with the structure. The use of Polyester Roundslings also increases the flexibility of the rigging, allowing highly stressed slings to stretch more than their lightly loaded neighbours. Additionally, I would include small lateral loads as a fraction of the lifted weight, to account for off-level situations.

When the crane hooks up to the load, they'll rig it based either on a rigging drawing (if available), or just based on where they guesstimate the Center of Gravity is. There is often a large variation from where the designer calculates the CoG and where it lands during actual lifting. The crane's hook needs to land immediately above the CoG. If that CoG isn't where the rigging puts the crane hook, the load picks off-level, imposing lateral loads on all components of the structure. This is the reason why most of my engineered lifts were designed based on a CoG calculated from load cell readings.

In short, I would model the load with standard load factors, being on spring supports, make them quite flexible, add some small lateral loads (2% - 5% of the lifted load, depending on how cowboy you think the erector will be), and add a small "cowboy factor" for unforeseen circumstances. That cowboy factor could be 1.0 for an engineered & supervised lift, higher for something happening out in the sticks with no supervision.

Just to get some visualization, here's a photo of a compressor skid being lifted by a mobile crane. It gives you an idea of the sheer amount of rigging and how much movement it could tolerate. There's also a good amount of cable in there, which has a good deal of flexibility.

 

[phamENG]

Craig H - based on your response I'd say you aren't late - we were here early.

Thanks. That's truly invaluable insight. I appreciate you taking the time to explain everything. It sounds like your advice lines up pretty nicely with the Australian documents posted by steveh49. The added thoughts on the pad eyes are good, too.