Bracing and gusset plate connection question

[faromic80]

I have some questions about gusset plate connections in general. They may pertain to vertical bracing or truss connections. I'm attaching a gusset plate to the web of a column and underside of a beam (the beam frames into the web of the column). The brace force is about 22 kips (from STAAD model). Some of my coworkers have said that they will typically double the brace force so that the lateral force resisting system is designed to take tension only. Anyways, let's say the brace is at 45 degrees and that the horizontal component is 31.1 kips (.707*44 kips). I'm wondering how others detail this connection; since the gusset plate is attached to the web of the column, do you provide vertical stiffeners at the center of the web opposite the gusset plate to aid in resisting the horizontal force? Maybe you use horizontal stiffeners? Vertical stiffeners seem more efficient in resisting the tension compression loads.

Another question about the connection design examples on the AISC 13 CD. In particular, example II.C-3 problem C. On page IIC-38 in the gusset to column connection box, they calculate H.ac as 0 kips. In the next box they calculate the horizontal force as the horizontal component of the brace force. I guess they're assuming the horizotal force will be resisted by the weld between the beam and gusset plate. If so, is this horizontal force not resisted by the vertical weld (column web/flange)? Is there a reason for the way they calculate this? A coworker of mine said something to me about how welds are designed with different allowables based on the direction in which they're loaded? I never heard of this nor did this. I always used .3*0.707*weld size*70 ksi.

Thanks

 

[nutte]

There are a lot of questions here...

If you want to do tension only bracing, model it this way, don't use T/C braces and double the force.

You need to study the Uniform Force Method. In particular, you could redistribute the forces so that you have just vertical shear at the column web. Then you have no horizontal force to worry about, and no need for stiffeners.

I'm not real sure how a column web would have vertical stiffeners...

Fillet weld strength does depend on the angle between the weld and the load. Look at J2.4 in the specification, specifically equation J2-5.

 

[SteelPE]

I believe the answer to your last question has to do with the way the problem is solved. They are using the Uniform Force Method to design the connection. From page 13-3 of the manual

The essence of the Uniform Force Method is to select the geometry of the connection so that moments do not exist on the three connection interfaces, i.e. gusset-to-beam, gusset-to-column, and beam-to-column. In the absence of moment, these connections may then be designed for shear and/or tension only, hence the origin of the name Uniform Force Method.

 

[JAE]

Usually we place horizontal stiffner plates, from flange to flange, to brace a web that has a gusset plate welded to it on the other side.

This detail is fine for moderately loaded connections. I'm not sure I'd use it for a full seismic application as the stiffner plate and the gusset plate are at 90 degrees from each other.

 

[faromic80]

I was also thinking horizontal stiffeners plates. I was just asking because a coworker of mine mentioned using vertical stiffeners which didn't seem reasonable to me. I've looked at the uniform force method a lot the last couple of days. my w.p. is located so that no moments exist.

nutte: What do you mean that there would be no horizontal force? it has to be taken out somehow, doesn't it? If the brace is at 45 where would the horizontal force go? Into the weld between the beam and gusset plate?

JAE: which detail are you referring to?
"This detail is fine for moderately loaded connections. I'm not sure I'd use it for a full seismic application as the stiffner plate and the gusset plate are at 90 degrees from each other."

 

[nutte]

You can redistribute forces delta V and delta H to get your interface forces/moments where you want them. You coud take the delta H through the gusset to beam weld, inducing a moment on this weld. Welds with moment are easy to check. Then you have only vertical shear at the gusset to column connection.

 

[JAE]

I was referring to the use of a horizontal stiffner on the back side of the column web.

You could also use a vertical stiffner in conjunction with a horizontal stiffner - the vert stiffner would line up with the gusset and be on the back side of the web and "span" from the column cap or base plate to the horizontal stiffner located in line with the end of the gusset plate.

 

[jike]

In a multi-story building the only force that passes thru the column is from the collector, otherwise the brace forces from above get transfer from the beam to the brace. The column, however, picks up the vertical component of the brace. The only exception to this is if there are multiply lines of bracing (or some bracing stops) where forces from above get redistributed at this level.

 

[faromic80]

JAE said:
"This detail is fine for moderately loaded connections. I'm not sure I'd use it for a full seismic application as the stiffner plate and the gusset plate are at 90 degrees from each other."

I have several questions so sorry in advance.

What would be an appropriate detail for seismic. why would you not want to use 90 degree angles like that.

I don't have seismic in my case. I guess another question I have is if horizontal stiffeners are really necessary for this application when attached to column web. In the case where the w.p. location doesn't create moments, there shouldn't be large horizontal force on the web and thus stiffeners wouldn't be necessary, right? Would you just check the web of the column as a point load on the web (the web spanning a distance T). To check the local stresses on the underside of the beam would you check shear in the flange (the area of consideration being the thickness of the gusset plate and thickness of the flange).

I also read in the AISC manual about transferring the vertical force from the beam to gusset connection to the column to gusset connection. This induces a moment in the plate and beam to gusset connection. How would you check the moment in the plate? Is the Sx the thickness and length of gusset plate?

 

[faromic80]

I'm back on this project again and still have the questions above. sorry for posting again but maybe this thread was missed. The only thing I'm wondering about is if you typically provide stiffeners at the opposite side of the web for local web effects. Technically I could design the gusset plate webs to resist shear only. In this case, there would be no tension on the web, only shear. then the stiffener really would be required, correct. So I guess it's really a matter of how you look at it.

 

[WillisV]

I do typically provide a horizontal stiffener at the opposite side of the web mid-height of the gusset just to stiffen the connection up and make it behave as analyzed.