Steel Tube Beam to Steel Column Connection

[NewbieStruct]

I have a situation where I am analyzing a steel tube roof beam sitting on steel tube column. The beam is welded to the column. This is an interior beam and is not part of any lateral resistance system per the existing drawings. We are adding a new mechanical unit on the roof. If I design the beam with existing loads, it fails in the deflection criteria. I am assuming the connection to be simply supported. If I assume the connection to be fixed, it is okay in deflection. Is assuming the connection to be fixed correct since it is welded?

 

[JAE]

Not fixed since the column will have some flexibility.

I would model the column as well to take into account the "partial" fixity there.

 

[TXStructural]

You say that the beam sits on the column and is welded to it. Is it a single span? The relative stiffness of the members does come into play. If the in-plane stiffness of the columns is similar or greater than that of the beam it is fixed, otherwise model the whole frame with the beam-column connection fixed.

 

[NewbieStruct]

I just visited the site to take a look at the beams and columns. It is a two span beam continuous over the column. There is no ceiling throughout the building. When I design beam assuming there are no ceilings, using the L/180 for live load and L/120 for total load per the IBC, the beam does satisfy the deflection criteria. Is it OK to use that criteria since there are no ceilings. If I were to design a building on my own and the building had no ceiling should I use this criteria. I have generally use L/360 and L/240.

 

[connectegr]

If I picture your description correctly... three columns; one at each end of the beam and one at the center span of the beam. The beam bears on the columns at the three supports, and is welded to the columns. Thus two spans...

I would consider this pinned at the outside supports and fixed at the center support. A continuous member is the same as two fixed ends at the same support.

http://www.FerrellEngineering.com

Providing fabrication and erection efficient structural design of connections. Consulting services for structural welding and bolting.

 

[Ron]

Deflection is a serviceability criterion, not a failure criterion. If your serviceability will allow L/120 then that works if the beam is not overstressed. Ceiling deflection criteria were established when ceilings were directly connected to the structure and were generally rigid (plaster) or similar. Since the advent of suspended ceilings, those criteria do not apply.

 

[wannabeSE]

Let me jump on a soap box. Often times, the deflection limits in the IBC are too heavily relied upon. Engineering judgment and coordination with architectural and other engineering disciplines are more important. As Ron mentioned, deflection is a serviceability criterion. An engineer needs to understand what is in the building when deciding deflection limits. L/240 live load deflection limits are not adequate on long spans when the partition's top of wall is only good for 3/4" deflection. This stuff need to be coordinated. There is a proposed amendment to the 2015 IBC that will clarify the intent of the dead load deflections. The dead load deflection criteria is meant to account for long term creep. See page 200 of these proposed amendments:

http://www.iccsafe.org/cs/codes/Documents/2012-13cycle/Proposed-A/02_IBC-S.pdf

For the question at hand, ensuring the roof does not pond is more important than insuring the D+L deflection is less than L/120. Footnote g in IBC table 1604.3 allows you to take the dead load as 0 with steel construction because it does not creep. But large dead load deflection may be noticed and the general public may think that the structure is overloaded or failing (and their perception is important).

 

[kylesito]

wannabeSE nailed what I was going to add. Ensuring that you don't create a ponding situation is more important than the ratio.

Just an idea, often when we add RTU's to a structure post-construction, we stub up directly from the columns and build an independent frame. This eliminates any contribution of our system to the roof and transfers everything directly to the columns.



PE, SE
Eastern United States

"If a builder builds a house for someone, and does not construct it properly, and the house which he built falls in and kills its owner, then that builder shall be put to death!"
~Code of Hammurabi

 

[NewbieStruct]

Thank you for your input. We are planning to strengthen the beam since the mechanical unit is heavy. I will make sure that the deflection is reduced too.