I-Beam vs. Square Tube for Columns

[DRoam]

Hi, we're in the process of designing a single-story mezzanine which will support 3-4 components weighing up to 10,000 lb. each. I've attached a picture showing the framework we're using.

My question is regarding the columns. My supervisors want to use I-beam for the columns. But everything I'm seeing from doing a quick search on mezzanines shows structural square tubing being used for the columns.

So my questions are:

1) What is the reasoning behind using square tubing as opposed to I-beam? It's my experience that I-beams are generally used for the columns in building frameworks. So what is the motivation behind opting for square tubing for mezzanine columns?
2) Is there any compelling reason that we shouldn't use I-beam for our columns?

I'm a newly-graduated mechanical engineer and this is outside my experience and expertise, so any tips and wisdom are appreciated. Thanks in advance!

 

[jayrod12]

For small structures such as the one you're describing we pretty much only use tubes. However that is because in our area they are readily available.

 

[boo1]

round tube are often used

 

[spats]

Because the columns are most often exposed, a lot of times it's mostly for looks and cleanliness. The connections can be made to look a lot cleaner too.

 

[KootK]

This was a good, recent discussion of this issue: Link

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[BamaStrucPESE]

Is this an interior or exterior structure? Interior tubes might be ok, but exterior you are likely to get moisture on the inside of the tube and it could slowly corrode from the inside out without your knowledge.

 

[jayrod12]

Any exterior tubes should be capped with a fully welded steel plate. Then the corrosion issue is solved.

 

[RPMG]

W-Shape: Good because it is cheaper and higher strength material. Bad when too small, connecting to the web due to access, and the minor axis governs the design. The minimum section that we use for lightweight structures is typically W6x15. Beams connecting to the web have the bottom flange coped for constructability.

Tube Steel: Good because it has 2 major axes, connections are similar on both sides, and it can be very small. Bad because it's more expensive, and you're limited on connection type. Also, there's the exterior problems associated with tube steel. The minimum section that we use for lightweight structures is a pipe 4". Connections are limited to shear tabs and continuous beam connections.

More importantly, I think that there is a stability problem here along the 3-column direction.

 

[structSU10]

To what RPMG is saying, you probably want your beams in the 3 column direction to be as deep as the beams running between columns in the other direction, and to provide knee braces at each corner in each direction at least. Of course, it should be analyzed for the lateral loads and such as well as the gravity loads.

 

[DRoam]

Thanks all for the responses. I'm still weighing everything that's been said and doing some research to better understand these unfamiliar concepts.

In the meantime, some specific questions:
1) What is a "moment connection" or "moment frame"? Aren't most beam end connections capable of carrying a moment?

2) @KootK, in the post you linked to, what is meant by a "knife plate connection"? I did a quick search and didn't come up with super clear results.
3) @KootK, also in that post, what do you mean by "non-orthogonal load cases"?

4) @RPMG, what is the stability problem? Is it that there's no bracing against the structure 'leaning' along the 3-column direction?

5) Can any of you recommend a really good 'fundamentals of structural design' book or PDF that will be a good reference for someone like me, who just needs to know enough about structural design to design safe, reasonable, simple structures, similar to the one I attached above?

Thanks again for all the input!

 

[dukengnr79]

You should probably also look into adding stiffener plates at those locations where the continuous beams run over the columns (if the compression flange starts to displace relative to the top flange of the beams you can lose stability of the column), especially if you have some heavy components up there on the roof.

Perhaps using deeper beams as mentioned above in both directions would address this situation by stiffening the web running over the column, and you could use a stiffener plate at the ends of the canopy which have only one beam framing into the girder above the column.

 

[mike20793]

1. Briefly, a moment connection is stiff enough to not allow rotation at the connection. Typical connections are designed for shear only because of the amount of rotation. Moment connections require special detailing.

2. A knife plate connection uses a single plate to connect to a member.

3. Non-orthogonal load cases apply load to the structure in two orthogonal directions at once to simulate load not being applied perfectly normal to the building faces.

4. The way it's framed is difficult to resist lateral load in the 3 column direction.

5. The good design books are material specific. There has been some good threads recently about the better ones.

 

[KootK]

1) What is a "moment connection" or "moment frame"? Aren't most beam end connections capable of carrying a moment?

Moment connections are connections capable of transferring moment. Not all are. In fact, most common connections are considered incapable of serious moment transfer.

) @KootK, in the post you linked to, what is meant by a "knife plate connection"? I did a quick search and didn't come up with super clear results.

Google "single plate shear connection" or "shear tab". It's just a vertical plate welded to the side of a column that the beam web gets bolted to.

3) @KootK, also in that post, what do you mean by "non-orthogonal load cases"?

Sometimes, wind and seismic loads must be considered that are directed at an angle to the primary axis of the building. These are non-orthogonal load cases because they are not aligned with the primary orthogonal axes that most buildings possess.

5) Can any of you recommend a really good 'fundamentals of structural design' book or PDF that will be a good reference for someone like me, who just needs to know enough about structural design to design safe, reasonable, simple structures, similar to the one I attached above?

Really good? Eh. Any single reference is bound to be a mile wide and an inch deep. These are the best that I can think of for your purposes:

Link.
Link

Structural engineering is a complex animal. That being said, mechanical engineers seem to do exceedingly well at it.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[RPMG]

A moment connection connects the beam/col flanges. In the 2-column direction, the "flanges" of your pipe are connected to the beam, so this is a moment connection within a moment frame.

A shear connection is only to the beam web, which acts like a hinge. This is the case for your minor beam ends in the 3-column direction. This is the stability problem. If I were to displace the top 1", it would form a parallelogram, and the structure lacks something that will intentionally force it back to its original shape.

I would recommend that you add knee braces at the central posts (each side) as a minimum.

 

[DRoam]

Thanks all for the input.

I still feel like I'm doing something very outside my expertise and I'm uncomfortable making these design decisions. So I really appreciate all of your help. I have two more questions.

1) As it looks like we'll be using I-beam for columns, what type of connection should I use for a continuous I-beam running over the end of an I-beam column, as shown in connection 1 in the image above? I did a lot of looking and couldn't find many examples of that type of connection. The end of the girder is essentially "cantilevered", so I'll need some way of dealing with that.

2) How does the fact that the girder is cantilevered affect the selection process for the girder size? I'm using the AISC Manual of Steel Construction: Allowable Stress Design to size the beams based on the load we'll have on this mezzanine, and I understand the use of the beam tables fairly well, but I'm unsure of how to deal with the effect of the cantilever. We WILL be putting a considerable load on the edge of the mezzanine which is supported only by the cantilevered girders, as much as 1500 lb distributed along the distance between the girders.

Thanks again for all the help.

 

[jayrod12]

1) Cap plate on the column with bolts through the beam flange for the connection at 1.

2) It only affects the girder size in the sense that you need to meet strength and deflection requirements for the cantilevered portion as well as the backspan portion. Don't forget to look at patterned loading for worst case scenarios.

However, the detailing will require more attention than choosing a beam size. There are a few threads around here recently about bracing beams and the discussion about bracing cantilevers is brought up in almost every one.

 

[dukengnr79]

DRoam, did you also decide to go with some sort of knee bracing in the long direction of the canopy?

 

[DRoam]

@jayrod12:

1) Should the cap plate be rigidly attached to the column's web AND flanges, or only to the column's web, as depicted here? Link. I would guess the reason for only welding to the web would be to prevent it from becoming a 'moment connection' and transferring moment to the column. I would guess this is what we want. But to me, looking at that connection, it looks like too much bending in the beam could tilt the cap plate, thereby trying to pry the cap plate away from the column web--in other words, 'compressing' the weld on one side of the web and trying to tear away the weld on the other. Is this a possibility/concern?

2) So the beam sizing design doesn't take into consideration the fact the fact that I have a load on both sides of the vertical support? It seems like this would be an important factor. For example, while analyzing the backspan portion between the two columns, ignoring the load on the other side of the columns is ignoring the fact that there is additional moment being applied to the ends of the beam, as well as additional shear at the location of support.


@dekengnr79:

Yes, I did decide to brace the 3-column direction somehow, probably with knee bracing. Also I've been wanting to ask, could you explain what you meant by "if the compression flange starts to displace relative to the top flange of the beams you can lose stability of the column"? Do you mean the compression flange could displace laterally due to local buckling of the compression flange? If so, why would this cause instability in the column?

 

[TLHS]

Yeah, you really need a local structural engineer to look at this for you. You're making a good effort, but the worry is that you don't know what you don't know, so you're going to have blind spots here.