Friday, February 24, 2012

February 8th, 2012

Picture from

Rzeszow, Poland. A tower crane with the classic Potain Arrow on the counter jib suffered a structural failure between jib section #1 and 2, possibly exclusively on jib section #2. The cause of the failure is not noted. Fortunately no one was injured in the accident. I find it incredibly important to note that the operator was "sober" as noted in the article. (I find this amusing as if drunkenness would be normal or a likely cause of a crane accident as if he were driving on snow. Not that it's likely as a cause, but I suppose a lapse in judgement has been known to occur during alcohol consumption.)

This is an interesting accident. The pendent is fully connected still. Jib section #1 is still aloft. Yet when you look at the end of it, you can see that the chord splices have ripped, presumably at their welds or just adjacent to them. Chord construction is commonly large angle iron pieces welded together for form a square member. At the ends, the splice connections, be they male or female, are welded on. If the welds at this point were not looked at since they are not obviously noted as welds in a cursory glance, a strong load sitting on this splice could easily cause a fracture of the lower chord splice welds which would then rip the top chord, as it has done, and the trolley would be missing from these pictures as nothing would be present to hold it up to be visible. Additionally, note the cables (ropes) appear to be broken near the weld failures indicating trolley placement.

Certainly wind could have played a role, as well as side loading. Side loading is less likely to be a cause this far in on the boom as leverage is severely reduced and the slewing motors would have a difficult time causing this damage at a 10 meter radius. Wind is also commonly caused with the trolley farther out. See this South African accident as an example.

I feel confident that this was simply a structural failure. We have an older model of crane and I would ask, who is looking at this crane? When they inspected the boom, did they look at those welds, especially underneath the chords? Normally these jibs are flat on the ground and a mirror would be necessary to do this. Inspect your welds regularly. I have my jib walked monthly at a minimum. We spend an hour going through the jib, trolley and ropes. Obviously I look that the easily reachable items daily. You too should take the time to look over each weld. You'll find nothing 99 times, but that 100th time you may save your own life.

Friday, February 17, 2012

February 16th, 2012

New York City, NY Imagine a empty dump truck falling from the sky at near terminal velocity. It virtually happened. One of the Favelle Favco tower cranes at the World Trade Center suffered a load line failure while hoisting 3 girders weighing up to 40,000 lbs around 40 stories. The load came crashing back down on to the truck it had been hoisted from. No one was reported as injured.

For me, this is a mind blowing accident. Favco's can hoist 40,000 lbs on a single line pull, depending on the model. Not only that, but at a rate of 190 meters per minute. I'm on a pretty short crane right now because we are on a two crane site, but I'm feeling pretty certain that even with a empty hook, it doesn't move at 190 meters per minute, let alone with it's maximum load. The point being, it's a horse. Both a Clydesdale and a Quarter horse in one. So what could the cause be? is reporting that the crane had a rope rubbing a bolt in October of 2011. Simple rubbing is not a cause for replacement. Flattening of individual wires to 2/3 their original diameter is cause for counting the wire as broken. Crown breaks would be another cause to count them as broken and then the rejection criteria according to manufacturer of the crane and the rope would begin to be sorted out. So did the company replacing the cable not deal with the rub problem root cause? Maybe, but not likely. The crane would have been inspected since then, and likely beyond the daily operator inspection. Not that this rules out error.

Another cause that occurs to me is the potential for the wrong rope to be installed. Tower crane load lines are to generally designed at a 5:1 ratio. So if you have a 40,000lb capacity on a single line pull, you need a 200,000 capacity rope. What strikes me is the potential for a incorrect rope installation. While this would be stunning, would it be more stunning than not properly addressing a bolt rub? Think about it, if you have a non-rotating rope, you have a swivel at the attachment point. If you have a rotating rope, you have a fixed attachment point. You can also have right hand or left hand lays depending on the drum rotation and attachment side that can affect the design and installation. Were one of these detail missed in the October rope replacement? I was talking with a friend about it this morning and he wondered about a rope defect. After some consideration, I thought, If you have a 37x7 rope, you have 259 individual wires. What are the odds that you would have multiple failures within one rope? While it can't be ruled out, it doesn't seem likely. I have to lean toward the wrong rope being installed. You can't have left a potential known rub point as a problem... right?

The lesson for us is daily inspection. Don't show up early and take a morning nap in the cab. Go through the crane. Look at the rope on the drum. It's not complete, but it's better than putting your feet up and setting the alarm. Make sure that we are going through our ropes against the horizon regularly and monthly have the bell man wrap it with a towel as we hoist up or down as appropriate. I've seen 9 month old ropes birdcaging off the drum with the hook at the lowest point. If I hadn't simply walked back to the drum, who knows what would have happened. Look at all of the rope. Make it a habit. Think of the ride this operator must have went through losing 40k and the fear that he might have just had a death under his hook Take that thought with you. It's what should be motivating you daily because you never know unless you inspect.