U factor confusions

[Keystation]

Hi, I wanted to ask the community about my confusion on steel connections. For your reference, I use AISC.

1. Is it possible to have multiple U factor for the same tension members? For instance, a diagonal member will experience tension in two directions. Do you calculate U factor when the load is x dir and y dir separately? Do you use the area of the connection in that direction (x or y)?

2. For angle connections, say both legs are bolted, do you use gross area of the whole section to calculate tension or do you use just one leg that is on the load path. I have seen both. Where people using gross area subtracted by eff bolts hole areas or where only one leg is considered for checks. How do you know which area to use? Does this have to do with eccentricity?

2. Why is U factor applied for tension members. Are there any other scenarios where U factor is applied? U factor accounts for non uniform stress distribution. Wouldn't this happen the same way for shear, compression or flexure whenever the loading is eccentric? How is eccentricity covered in each loading scenarios? After all U is called "shear" lag factor.

Thank you for reading my unorganized thoughts! I had these confusion for a while and wanted to hear other engineer's opinion on them.

 

[cliff234]

My answers to your questions:

1. I never heard of multiple U factors for a connection. How can a tension member be loaded in more than one direction?
2. When you have a single angle tension member with bolts in both legs, there is no shear lag. U=1
3. Shear lag applies to tension members because when connections (bolts or welds) are not provided in each of the cross-sectional elements of a member, there is a lag in the ability of loads to jump out of one cross-sectional element without bolts or welds into an adjacent element with bolts or welds. (For example, when you have a single angle tension member with bolts in only one leg of the angle, the load in the angle leg without bolts cannot instantaneously jump into the adjacent leg with bolts. The shear lag factor accounts for the detrimental effect of the load having to jump into the adjacent leg that has bolts.)

 

[Keystation]

My answers to your questions:

1. I never heard of multiple U factors for a connection. How can a tension member be loaded in more than one direction?
2. When you have a single angle tension member with bolts in both legs, there is no shear lag. U=1
3. Shear lag applies to tension members because when connections (bolts or welds) are not provided in each of the cross-sectional elements of a member, there is a lag in the ability of loads to jump out of one cross-sectional element without bolts or welds into an adjacent element with bolts or welds. (For example, when you have a single angle tension member with bolts in only one leg of the angle, the load in the angle leg without bolts cannot instantaneously jump into the adjacent leg with bolts. The shear lag factor accounts for the detrimental effect of the load having to jump into the adjacent leg that has bolts.)

Thank you for the response! That makes sense. U can just be 1. I thought U had to do with specific eccentricity.

Based on your shear lag description, U accounts for unintended loading on other connections. If gross area were to be used for tension or shear, it would assume all area is acting uniformly without jumps. Do you use area of only one leg for connection design?