Moment connection on the web of steel column 2

[Enhineyero]

Hi everyone, I would like to consult regarding this: I have a two span steel beam (WF) with one span as a cantilever. My columns are all I-section. The beam is connected to the column's web. The moment of the cantilever beam is balanced by the adjacent beam (same section with the cantilever), such that the moment taken by the columns is very small.

After designing the req'd fillet weld thickness (assuming that the steel sections are thicker than the weld), what are the other things that would I need to check? is there any? since the load path is almost direct for one beam to another.

I attach some sketch for better understanding.

regards,

 

[renderu]

The column will share some of the moment. Unless you've proven that it doesn't?

 

[BAretired]

With a 15' cantilever loaded and a 15' backspan unloaded, it is clear that Point B will rotate, so the column will take moment in accordance with its stiffness.

Why do you want to splice the beam at Point B? Why not run it straight through, 30' long? Then rotate the column 90^o and provide a proper moment splice at Point B with stiffeners in the beam at the column flange locations.

BA

 

[BAretired]

Of course, if your main goal is to ensure the column takes no moment, then use a continuous beam with a rocker bearing on the column.

BA

 

[Enhineyero]

Thank you for your replies.

BA - The column will take a small amount of moment (about 10% of the moment of the cantilever beam) due to the low stiffness of the column with respect to the direction of bending. Ideally i would like to make the 30' long continuous beam, problem is that my column will not stop at this level and will continuous to support another level. Hence a need to cut the beam at pt B.

I could make a better moment connection at pt. B by rotating the column by 90deg but due to architectural requirement (its a pain) i am not permitted to do so. any other suggestion how i go about this?

 

[hokie66]

Assuming your beam will only see gravity loading, not uplift, I would use a bolted moment connection as follows: Use an end plate on both beams and splice plates for the top flanges. The splice plate would have a complete penetration weld to the column web both sides.

 

[wannabeSE]

Ehineyero,
Is there a reason for welding the beam flanges to the column webs in lieu of a more typical moment connection to column web. Often times, a typical shear connection is used for the shear and horizontal connection plates are welded to the column flanges (similar to stiffener plates and sometimes welded to the column web) that are connected to the beam flanges for the moment. If you are in the USA, look at the Steel Construction Manual chapter 12 (in the 13the edition). I did not provide a very good description of the connection. If you don't have the Steel Construction Manual, let me know and I will try to give a better description connection and summarize some of the content in the manual.

 

[Enhineyero]

Hokie - Can you give a sketch of your idea? I cant visualize how to weld the splice plate to the web of the column if there is an end plate on the beam.

WannabeSE - My intention is for the flange to transfer the force to the weld, then the weld will transfer it to the web, then the web into the flange of the adjacent beam. Since both beam are of the same size, the force transfer is almost direct. Actually i am using BS code (which is similar to AISC code in terms of principles with regards to design).

My understanding is that the horizontal stiffener plates that connects the column flange and the beam flange is used to "engage" the column flanges to the bending action that is produced by the beam, otherwise bending might be localized to column web, I am correct? Since the moment of the cantilever beam is balanced by the connection of the backspan (meaning column will take a very small unbalanced moment) do i still need those horizontal stiffener plates?

 

[renderu]

You checked the column for buckling after that moment is applied?

 

[hokie66]

Enhineyero,
As my proposed detail is quite simple, I will persist with explaining my "word picture" rather than bothering with a sketch. The end plate would start at or below the top flange, so doesn't interfere with the splice plate. I would want 4 bolts in the end plate centred about the bottom flange, and additional bolts if required for shear.

Rather than use complete penetration welds on the splice plates, it may be preferable (and better) to slot the plate through the column web and just fillet weld it.

I wouldn't be concerned about moment being taken by the column. The beam has to be strong enough to take the propped cantilever moment, so any moment the column absorbs about its weak axis is inconsequential.

 

[BAretired]

It is much easier to splice the column instead of the beam. Run the beam through continuously with a fitted stiffener between the webs of the columns for bearing. If more bearing is needed, use a pair of channels, same depth as column back to back.

Place a cap plate over the lower column and a base plate under the upper column, each with four bolts through the beam flange. A small moment will be transferred to each column but that will be true in any case.


BA

 

[dhengr]

I would be inclined to go along with BA’s approach because it seems a little less complicated and it keeps the most important member at that joint continuous as it should be. We don’t know what the loads or member sizes are so the detailing could vary a little. My stiffeners through the depth of the canti. beam, inside to inside of flanges and to the web, would be a split Tee made out of a piece of the column section, and fitted and welded btwn. the flanges and to the web.

 

[BAretired]

I like dhengr's idea for a stiffener.

BA

 

[hokie66]

Respectfully, I disagree with BA's approach in this case, and believe my way is simpler. I like my columns to be in as few pieces as possible.

 

[dhengr]

Hokie:
I like columns in few pieces also, so no disagreement there. Except for this crazy detail where we have to make some determination about which is the critical member to keep continuous and as clean as possible. As I understand your scheme, you would slot the col. web to pass your splice pl. through the web. Then you would shop weld the splice pl. to the col. web. This weld better be well detailed because the slot removes 10-20% of the col. mat’l. in compression, and potentially leaves a nasty slot end notch w.r.t. col. moments. Then, you have two end plates on the beams and bolt through the col. web to complete your connection for shear and any uplift on the canti. end. Finally, you fillet weld the splice pl., downhand, to the two top flgs., and that sounds doable too. Except, the first few inches of those fillets are very highly stressed as they start to pick up their load, something well above the avg. fillet stress. They are also in an inferior fatigue class/group as a detail for the first 6" or so. Then in addition, if there is any slip in your bolted end pl. connection, you will literally apply an unzipping force to the tips of the splice pl. fillet welds, as the beam settles. That’s a nasty detail however you cut it.

 

[nutte]

Assuming the required moment was not all the way at the beam's plastic capacity, I would weld flange plates to the inside of the column, and bolt the beam flanges to those plates. Use a shear tab to the beam web. This is shown in Figure 12-12(b) in the 13th edition Steel Construction Manual.

 

[hokie66]

dhengr,
No, my top flange connection would be bolted. Similar to nutte's solution, but rather than using a flange plate at the bottom, which makes the beam difficult to insert, I just use an end plate. Simple, not nasty. I have used it several times when a column is oriented perpendicular to a beam. The only weldments are the top flange plate to column and partial depth end plates to beams.

 

[Enhineyero]

Thank you everyone for your suggestions. I decided yesterday that I would make the beam as one cont. 30' beam. sitting on a cap plate of the column. I decrease the column size above so it could be welded to the beam (I will add stiffener plates on the beam on pt B). The loads that will be taken by column above would be very light, and the moment of the cantilever is much more critical compared to the load that will be transferred by the column above.

If the column above were carrying a heavy load then i would have done a different detail.

regards,

 

[hokie66]

Enhineyero,
Why the propensity for so much site welding? Most of us prefer doing the welding in the shop and then bolting in the field.