Australia: overhead sign fell onto car

OMG :eek:
Thats not featherweight, so very deadly.
 
Bit surprised a sign that size up in the strong winds was only held at the bottom. All the others are probably suffering metal fatigue as well!
 
Looks like it took a chunk of the horizontal beam with it, so either faulty manufacturing, bad welds, or a combination of the 2. Scary stuff!

Every sign that big here is mounted at both top and bottom, or is on breakaway poles on the side of the road.
 
I an thinking use of the wrong strength bolts, maybe sub par Chinese bolts.

I would think you need at least 8.8 bolts, maybe 12.9
 
Looks like it took a chunk of the horizontal beam with it, so either faulty manufacturing, bad welds, or a combination of the 2. Scary stuff!

Every sign that big here is mounted at both top and bottom, or is on breakaway poles on the side of the road.
Wind speed increases with height, up there it may regularly get 100mph gusts and with no support at the top there will be a huge amount of torque on the bottom mount trying to twist it off, probably a weld fractured, or the metal surrounding the weld suffered metal fatigue. With a top and bottom support there would be zero torque! The poles at the side of the road are normally sunk into the ground so are supported over an meter or two height removing the torque on a single point, there is also a lot less wind speed down near the ground.
I an thinking use of the wrong strength bolts, maybe sub par Chinese bolts.

I would think you need at least 8.8 bolts, maybe 12.9
Bolts would be being used as they are designed to be used so probably not a bolt failure.
 
The metal plate sign face is stiffened by a series of horizontal and vertical ribs, which are secured to a structural frame of square hollow section steel members. There would be no reason to secure the sign at the top. It's just like a big garden fence panel fixed to strong posts that are embedded in the ground or bolted to the floor.

1547201279261.png

I understand that this steel frame was secured to the supporting gantry by both welded and bolted connections, which is unusual. In my experience you would use one or the other, but not both. In the photo below you can see the 4No bolts in the corners of each end plate, and the smaller square section that may have also been welded to the gantry although that appears to be rusty. The steel frame end plates were bolted to short supporting stools. These stools were welded to the gantry box structure.

1547200599811.png

My initial thought is that bolt fracture is the likely cause of the incident, leading to lack of restraint and allowing the sign to fall. Having seen further video evidence I now think weld failure is more likely. The bolted connection appears to be intact.
 
Last edited:
Could be faulty hardware, either bad QC or counterfeits. Someone could have used the wrong grade of hardware or improperly torqued them. The article did mention they checked the bolts on similar signs and found them OK. I'd think a design with less leverage better, but there's also failure mode to be considered- would it be better to lose just the sign or the entire structure to winds? I think the design needs improvement though; in a high-wind area you allow for those added stresses as well as the human factor of less-than-perfect installation and maintenance. Lots of torque stresses in any end-mounting schemes be they signs or fences, and the usual failure point is at that interface, so anything which reduced that torque loading would help.

Phil
 
Presumably the brown stuff is a large amount of sand, left by the wind? The bolts all look OK, and the welds appear to be painted over, so I don't understand how it was attached?
 
The video in the original post has been updated. New clips show that short steel stools were welded directly to the gantry. The sign face structure was then bolted down to these stools, which would allow for easy sign replacement and is quite typical construction.

The images below show two holes in the top of the gantry box where the stools used to be, and up close it appears that the welded connection failed. On this evidence I no longer think bolt failure was the cause as the bolted section is still intact.

I will endeavour to locate a structural engineering assessment of the incident in due course.


1547203661467.png

1547203702704.png

1547203752182.png
 
Probably metal fatigue failure or over-stress of the box girder thickness. Gusseting would have spread the bending loads over a larger area preventing this type of failure, especially if they extended across the full width. That would utilise the entire box section for strength, not just one forward face allowing the top face to flex. Needs a redesign upgrade as it's going to happen to all similar signs of the same vintage in that area :eek:

Phil
 
No Weld of mine ever failed, but i have made repairs i had to make a note of needing to be redone ASAP.
For instance a chemical tanker i sailed on, the pump assembly that was down in the stainless steel tank just where the pipe pick up the highly flammable / explosive stuff, had shaken loose and snapped several 15 mm thick stainless steel support brackets, and all i had for welding in stainless was arch welding electrodes with a 1.5 mm core, where for a job like that a 3.25 mm core would have been much more it.
But i managed to get the things to stick together in spite of it was actually only my #2 time welding in stainless steel with a arc welder, and the #1 time was a travesty so i had to get a more experienced welder to do it for me as it was just 2 pieces of thin sheet needing to be tacked together.
I do like arc welding, and was pretty good at it as a apprentice, TIG / MIG & MAG is okay too as well as oxy / acetylene welding, just not as challenging IMO.
As a apprentice on trade school we had a 2 weeks of arc and oxy / acetylene welding, i did all my work in 2 days, TIG / MIG & MAG was not on the books at all, bud i did do a lot of MAG on trains.

A good looking welding can almost make me exited :)
 
One other thing I've seen in big signs here that are installed in high wind areas are perforated or open grid (similar to a metal grid you'd see in a big ship or factory in an overhead walkway) so there's much less stress on the entire structure. You see it a lot on the gulf coast, where hurricanes are a regular occurrence. Strangely, I couldn't find a picture of a sign made that way. My Google-fu failed me. At least I found a picture of the flooring I was talking about. I'm sure making signs this way is a little more expensive than just sheet metal with paint or stickers, but they also last a lot longer in places where it storms a lot or is just really windy. Only downside is that the green flat area is not as reflective as a regular flat metal sign.

HTB1oohEJpXXXXbmXFXXq6xXFXXXw.jpg
 
One other thing I've seen in big signs here that are installed in high wind areas are perforated or open grid (similar to a metal grid you'd see in a big ship or factory in an overhead walkway) so there's much less stress on the entire structure. You see it a lot on the gulf coast, where hurricanes are a regular occurrence. Strangely, I couldn't find a picture of a sign made that way. My Google-fu failed me. At least I found a picture of the flooring I was talking about. I'm sure making signs this way is a little more expensive than just sheet metal with paint or stickers, but they also last a lot longer in places where it storms a lot or is just really windy. Only downside is that the green flat area is not as reflective as a regular flat metal sign.
Ours seem to just be kept short, made wide rather than high, so that the wind has less leverage. I wonder if someone just made the sign too tall for the design of mounting?

The big holy fences at the sides of the road are designed to slow the wind without causing turbulence so that large vehicles don't have to be banned in high winds - this is definitely a high wind area, I am above the sea traveling from England over to Wales:

1547216744858.png
1547216918386.png
 
Last edited:
Someone might have put a taller or larger sign on than had been engineered for. But regardless, a good base design would have had gussets extending to the rear face of the box girder simply to eliminate any possibility of the top surface flexing. The face side of the stub bracket had the full width of the face of the box girder to support it, but the rear of the stub bracket had only the wall sectiuon thickness of the top surface of the box girder supporting it, so naturally that top surface would flex or bend before the front surface did due to the disparity of support, You can clearly see where a section of the box girder tore away, indicating the weld held at least in that area. By adding triangular gussets to the side of the stub bracket, there would now be the entire width of the box girder resisting movement at the rear, same as the front. Thus you'd have the entire strength of the girder available instead of only a fraction of it as it is was.

That it failed within a year tells me some engineer needs to have his certification revoked when a simple carpenter like myself can see the flaw and understand it when they can't. I'd advise against driving on that road until they rectify this as fatigue failure could manifest itself with even a slight force once the metal has been overstressed.

Phil
 
Back
Top