Square Tube Column to H-Beam Connection 1

[BWally]

I have a “square tube column to H-beam” connection that consists of an H-beam resting on top of a square tube column. The beam ends at the column. There will be a base plate welded to the top of the tube and bolted to the bottom flange of the beam. There are no beam moments or beam axial forces of any significance to deal with here—only the shear at the end of the beam becoming a compressive axial force in the column. (The bolts will be arranged in such a way as to basically prevent this connection from having any moment capacity.)

How do I check the concentrated force capacity of this connection?

I’ve thought of a couple of logical methods of checking this, but logical doesn’t always mean valid:

1. Check the beam using the AISC ASD Chapter K concentrated force criteria that we all use for concentrated forces on beams. Use the square tube width as the bearing width. If the web is too weak, size a stiffener on each side of the web using the Chapter K guidelines. For the assumption that bearing width = tube width, the base plate has to be able to make this happen: it has to be able to get the load from all 4 walls of the tube into a length of beam equal to the tube width without becoming overstressed itself. How do I determine the required thickness of the base plate to make this happen?

2. Check the beam using the AISC ASD Chapter K concentrated force criteria, but use the square tube wall thickness as the bearing width for half the load, assuming only the 2 walls of the square tube perpendicular to the beam axis are active for the load transfer at the connection (ie, the tube becomes an H-beam, with only its “flanges” transferring the load; its other 2 walls are inactive). If the web is too weak, place a pair of stiffeners above each of the tube’s “flanges”. The base plate thickness is added to the thickness of the beam bottom flange to get “tf = base plate thickness + beam bottom flange thickness” for use in the Chapter K equations. To check the base plate for getting the load into the tube walls, I would use the AISC LRFD Spec for Steel HSS (yr 2000), which gives some equations for the situation of an axially loaded square tube with a base plate. There is a bearing width involved in checking the local stress in the square tube wall, for which I would use the beam web thickness. There is also a “t1” term, for which I would use “t1 = actual base plate thickness + beam bottom flange thickness”.

 

[dik]

I try to extend the beam over the column. The issue is not likely the beam but the column load capacity with the intended eccentricity... if I understand your question.

Dik

 

[BWally]

Dik,
The beam rests fully on the column. Loading is concentric.

 

[WillisV]

See p14-18 of 13th edition manual for a dicussion of HSS cap plates. If the load is absolutely concentric to the HSS (and therefore no bending in the cap plate), then I would just use a 1/2" thick plate and be done with it.

If you reeeeally feel the need to do calcs you could probably assume a yield line pattern in the plate and check it with yield line theory.

 

[dik]

Not familiar with the AISC, but I would typically put a 1/4" plate stiffener centred on the column as just a normal detail. If the load was significant, or if it were taking additional load from a column above, I would design the stiffener assembly for strength. Generally, with the stiffeners provided, the connection is capable of most loads. The crown plate on the column would typically be 1/2" thick and be welded all around to the HSS column. For HSS I usually weld all around since any flexure stresses the corners more. With W columns, I'd likely put a full length fillet weld along the back of the column and two fillet welds on the inside of the opposite flange.

Dik

 

[swmiami]

When the beam deflects, its end will rotate. The load will shift to interior edge of the column (unless you place a narrow strip of bearing plate at the center of the column). I would either use an eccetricity of half the tube size or treat the connection as a moment connection and calculate the moment based on the relative stiffnesses of the beam and column.

 

[twinnell]

How can you arrange the bolts to keep this from being a moment connection? This type of connection typically has 4 bolts through the bottom flange of the beam into the column cap plate. This is a moment connection. You should also put stiffeners each side of the beam web to brace the top flange of the beam (this is a stability requirement per AISC).

If you don't want a moment connection, then instead of running the beam over top of the column, frame the beam into the column using either a face welded shear tab or a single thru-plate through the column.

 

[ashpkt]

Mentioning the actual loads and shapes would help us give you better direction. Nevertheless...

Ref: "Load and Resistance Factor Design Specification for Steel Hollow Structural Sections", AISC, November 10, 2000

8.3. Concentrated Axial Force on the End of a Rectangular HSS with a Cap Plate
When a concentrated force acts on the end of an HSS with a cap plate and along the axis of the HSS, the design strength ?Rn shall be determined for each loadedwall as follows. The limit state of local wall yielding shall be checked for both tensile and compressive forces and the limit state of wall crippling shall be checked for compressive forces.

For local wall yielding,
? = 1.0
Rn = (5 t1 + N)Fy t ? B Fy t

For wall crippling,
? = 0.75
Rn = 0.80 t2 [1 + 3 (N/ B/2)(t/t1)^1.5] Sqrt[EFy(t1/t)]
where
t1 = thickness of cap plate, in. (mm)
N = bearing length of the load across the width of the HSS, in. (mm)

As for your connection..."twinnell" is correct but, you will need to check concentrated loads on the tube wall.

 

[BWally]

Willis V,
Yeah, I’m the kind of guy who likes his calculations! Nothing makes me more content than doing a legitimate calculation! I'm not yet good enough to make “judgment calls”, as my boss likes to say. I’m writing this from home, so I’ll check out your reference tomorrow at work.

swmiami,
Eccentricity of bearing due to beam deflection never occurred to me, but it sounds like a legitimate concern. I’ll have to confirm this, but I think my beam’s deflection is so small as to warrant neglecting any end rotation.

twinnell,
In my case, there will only be 2 bolts, and they will both be on the same side of the column. It seems to me this would have little to no moment capacity. Even if there were 4 bolts—2 on each side of the column—having them just going through the bottom flange of a beam doesn’t seem to me like a moment connection. Sure, there would be some small amount of moment capacity, but how would you quantify it? I agree with you that stiffeners would likely be needed for beam stability. In my case however, I have a smaller beam framing in at 90 deg to my main beam that provides end rotational restraint for my main beam.

Ashpkt,
Force is 400 kips. Beam is a W36x135. Column is a TS10x10x5/8. Your reference is a good one, one I mentioned in my original post. Thanks for taking the time to reproduce it here. I would also have to check my beam web and not just my column walls. And to check my beam web, what do I use for bearing distance “N”? Is it the depth of the column, which assumes all 4 tube walls are active, or just the wall thickness of the column, which assumes only perpendicular walls are active? If it’s the column depth, then the base plate has to be thick enough to make this happen, right?

Thank you all for your time so far.

 

[JKStruct]

I agree with twinnell -- the four bolt connection is a moment connection. In fact, the AISC HSS specification (black book) addresses this connection in a convenient design example that transfers both axial load and moment (I do not have the book with me now, but it is in there). The theory is that the the moment is resolved into a couple whereby the compression force acts on the wall of the HSS column and the tension is carried in the bolts. You have to check the thickness of the beam's bottom flange and the thickness of the cap plate against prying action, or calculate the increased tension in the bolts due to prying action and account for it. But you can quantify the capacity. In my office, we have used a continuous-beam-over-column system as a "moment frame" for the LFRS (note: in general, only for one story buildings that are wind controlled, and the maximum moments are usually only 40-45 ft-kips factored, which we can take care of through this connection). But I agree with the other posters, and definitely add stiffeners for beam stability.

 

[BWally]

JKStruct,
I guess I stand corrected--you can have a full-capacity moment connection with just bolts going through the beam bottom flange.

What is the AISC HSS specification? I've never heard of it. It sounds like something I should get my hands on.

 

[WillisV]

The HSS spec used to be a seperate specification document within the AISC manual seperated from the main specification. The same held true with single angles (they had their own seperate spec). With the 13th ed. the HSS spec has been rolled into the main spec so there is no seperate "HSS Spec" anymore.

 

[BWally]

Thanks WillisV. I looked briefly through the 13th ed. manual and couldn't find the design example that JKStruct describes. Does anybody know where in the 13th ed. manual I can find this example?

 

[WillisV]

All examples were moved out of the manual itself to the accompanying CD.

 

[BWally]

I've been looking through the CD a lot lately. In fact, I looked through it just last night, and couldn't find anything like what JKStruct describes above. I did find an example of an I-shaped beam resting on an HSS column, with the load being only a force and no moment, which is applicable to my situation and very helpful. But for the case of a force AND a moment, as described by JKStruct, I couldn't find anything.

 

[JKStruct]

Sorry for the response delay...

It is true, the 13th AISC integrated angles and tubes into the spec, and their design examples are in a separate CD. I have not perused the CD much, but I don't doubt this example is not in there. The AISC HSS Connections Manual has the example on page 5-15, which is the moment connection chapter. Hopefully you can find the book.