Steel Beam to Column Moment Connection 1

[JoanWill]

Hi,

I have this I section steel beam to I section steel column using bolted moment connection.
I am wondering what happens if this moment connection fails? For the structure itself, shear connection is okay, but for some reason it is designed as moment connection but don't have enough moment capacity.

 

[jayrod12]

What failure mode is the connection failing under?

Can you post a sketch of the connection as well?

 

[phamENG]

Also depends on where that moment is coming from.

If it's for gravity loading, then it will depend on the failure mode as jayrod suggested. If it's a failure in the connection itself, the point of highest moment will likely shift back to the beam mid-span, which it can probably take albeit with excessive deflection. If the beam itself failed, then you're probably hosed and the beam will likely collapse.

If it's from lateral loading, then it can be anything from excessive sway to a complete collapse. Depends on the source, magnitude, and duration of the loading as well as the residual capacity in the shear connection to transfer some moment.

 

[human909]

At the risk of repeating what jayrod and phamENG have said it very much depends on what you mean by "failure" and how it fails and also what other load paths are available.

Most connections will fail in a somewhat ductile manner. If this occurs then you get reduced stiffness and likely moment redistribution to the beam (if possible). After all, the only real design difference between flush bolted moment connection end plates and a 'flexible' end plates is the thickness of the plate.

Also if there is no alternative load path then you are shit out of luck, as shear capacity alone won't cut it. Eg a cantilever beam.

 

[JoanWill]

Thank you guys.

It is a typical 8-bolt moment connection.
It fails at the connection (end plate yield and bolt tension)

The structure is checked using shear connection and it is ok. It is more to do with what will happen after the connection fails? Can I say it will simply becomes a hinge so as long as the connection can still take the shear then it is ok?

 

[phamENG]

I don't think that would be accurate with an end plate moment connection. A flange bolted or welded moment connection with separate components for the shear transfer, probably...but I wouldn't want to count on the post-failure remnants of this to do anything reliably.

 

[human909]

It fails at the connection (end plate yield and bolt tension)

Bolt failure is a problem. However it in many cases it is quite unlikely that bolt failure will occur especially if the end plate is yielding and your drawing is approximately to scale. The interior bolts are unlikely to fail if there the structure works when the connection is modelled as a pinned conection. How are you checking the connection?

That said the wrong type of connection could experience bolt failure if there is significant rotation. This is a problem and the connection could fail catastrophically.

The structure is checked using shear connection and it is ok. It is more to do with what will happen after the connection fails? Can I say it will simply becomes a hinge so as long as the connection can still take the shear then it is ok?

Maybe, probably. I'm hesitant to say yes because we don't have the full picture. Most moment connections that fail in calcs become semi rigid before failure. You'll need to plug this behvaiour back into your analysis to give you a better picture of what is occuring.

You need to put the full picture together yourself. EG I could design a connection that is weak in moment but apply a moment to it. I could plug it into some software ore use some hand calcs and I'd get all sorts of failures because my input is a moment that cannot be resolved without failure. If I instead take the same connection and call it a shear connection then it will pass the software and hand calcs.

I don't think that would be accurate with an end plate moment connection. A flange bolted or welded moment connection with separate components for the shear transfer, probably...but I wouldn't want to count on the post-failure remnants of this to do anything reliably.

I would happily count on in many circumstances. For example remove the exterior bolts and call it a flexible end plate and it will likely meet the requirements of a flexible end plate. As shown below.

(I've been busy recently designing both flexible and moment HSS to HSS joints. So joint stiffness and behaviour has been on my mind recently)

 

[dik]

I'm not sure what the concern is. I've designed dozens (or more) of these and the moment is taken care of by adding bolts and or thickening the connection plate. I just make sure that prying action is taken care of. I consider the bottom bolts acting in shear. I don't like to take the connected part to maximum moment... I generally like to keep that below about 85%.

-----*****-----
So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[jayrod12]

I'm assuming in the OP's case, the structure is already erected.

Seems odd that there would be two simultaneously occurring failure modes. Is it plate yielding(less concerning) or bolt failure(more concerning).

If the latter, maybe it's as simple as swapping for a stronger bolt.

Plate yielding would essentially turn it into a shear connection due to the excess flexibility in my mind.

 

[dik]

In these environs, swapping for a stronger bolt can cause a long delay...

-----*****-----
So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[jayrod12]

Do you have a better solution to what I presume is an already erected structure.

Come on dik, let's provide options as well as cautions.

 

[JoshPlumSE]

Bolted End Plate Connections can exhibit good ductility in seismic events. End plate yield is a desirable failure method. But, bolts tension failure is NOT.
Have you considered using higher strength bolts to replace the existing ones?

If it fails via end plate yielding, I would expect this to be a ductile failure mode. Meaning that moment re-distribution to other parts of the structure is the probable post-failure behavior.

 

[jayrod12]

As usual, Josh said things more eloquently than I.

 

[SWComposites]

Where exactly is the end plate yielding?

If there are bolts that truly fail in tension, then take them out if the shear check and rerun it.

 

[dik]

The option was in a post prior to that...

"...and the moment is taken care of by adding bolts and or thickening the connection plate."

and A490s are not common... they can be obtained, but may require a good lead time... Sorry guy... the comment's correct for these environs.

-----*****-----
So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[jayrod12]

Is thickening a existing plate reasonable?

I don't work directly with the guys as much as you do. But I feel that ordering new bolts even with a delay, is likely more realistic and construction friendly than cutting off the end plates, trimming the beam shorter, and welding on a new plate.

But maybe that's just my own ignorance.

I'll admit, if both are failing at exactly the same time, and it must remain a full moment connection, then it would make sense to change only the plates and add more standard bolts.