Braced Frame Connections

[Struct1007]

In the AISC Seismic Design Manual design examples for SCBF I am not understanding the f peak and f avg on the welds. At the chevron gusset plate, figure 3-13, the weld has shear,axial and bending. In determing the average, the axial and bending are conbined. The bending has a + or - sign associated with it, but because the axial is very small and the quantity fa +/- fb is squared, the signs are lost and the average is equal to the peak. Later when they are checking the gusset to beam or column connection in figure 3-16, they state that because there is no bending, the peak is the average. This seems to be incorrect at the chevron connection in figure 3-13. Can anyone enlighten me on this seemingly discrepancy?

 

[EngineeringAdam]

I am confused on your question - are you wondering why the avg equals peak? Or why there isn't bending within part of the connection?

I am not sure which of those you disagree with - please clarify.

 

[wannabeSE]

The Uniform Force Method is used to analyze the gusset at the beam to column connection. The benefit of using this method is there are no moments at the connections.

Have you looked at the Steel Construction Manual Part 13 on brace connections? Also, there are brace example problems for the Steel Construction Manual that can be downloaded from AISC.

 

[EngineeringAdam]

WannabeSE.

I am not sure that the AISC Seismic provision example the Struct1007 is talking about uses the Uniform Force method. When I rechecked the example there are indeed moments at the connections - which is correct. There are areas of the connection that don't have moments - like between the gusset and column, but this is because the forces here are coupled and can resolve the moment in the column without it impacting the gusset. There are moments between the gusset and beam due to unbalanced forces, which would be non-existent if the uniform force method was being used.

Struct1007 - please clarify your issue when you get time.

 

[Struct1007]

The question is in regards to the connection at the middle of the beam. The uniform force method is not being used. There is shear and bending on the weld and a very small amount of axial. The question I have is why is the peak equal to the average when there is shear and bending on the connection in figure 3-13 and the calculations on the next page ( 3-59 in my book )?

 

[EngineeringAdam]

Perfect, thanks Struct1007.

The example you are referring to does in fact come up with different values for both the Favg and Fpeak, it's is just the values are so similar that the answer was rounded to 16.1kips/in.

If you check the calculations by hand you will find that...

Fpeak = 16.1158 kips/in

and

Favg = 16.06 kips/in.


You see that they aren't exactly the same, but very close. This doesn't happen unless you Faxial stress is very low.

That make sense?

You pretty much nailed it by noticing that axial stress is so low - and that is your answer why the avg and peak are so similar in value. It is purely a mathematical issue.

 

[Struct1007]

One last question. Maybe I am over thinking this one. Lets ignore the axial for a moment. Shear is constant at 12.7. Bending varies and is 9.83 at ends and 0 in the middle. Stress at ends with SRSS is 16.06. Stress at middle is 12.7. Seems like the average is somewhere inbetween unless the average that they are speaking of is the average at the ends. Is it the average at the ends that they are talking about?

 

[EngineeringAdam]

I see exactly what you are saying, and I am not sure if I know the answer exactly.

As far as I understand it the elastic method for determining weld stresses isn't exact. It is taking the extreme fiber bending stress of the weld and assuming it is constant over the full weld - as you noticed.