Web opening in built up steel beams

[smjohns78]

Sorry if this is posted elsewhere, but I have a question regarding AISC DG2. This references seems to be written for doubly symmetric wide flange shapes. The term for the eccentricity of the hole, "e", seems to be based on the distance from the neutral axis of the nonperforated beam. If I had a beam with unequal flanges, common for composite construction, does anyone know if the term "e" would be based on the neutral axis of the non-composite section or should it be the depth of the section divided by 2?

Thanks

 

[dhengr]

Smjohns78:
I don’t have the benefit of AISC DG2 sitting on my desk, but I’ll take a stab at this. The potential problems with a hole in the web of a beam have to do with the types of, magnitudes of and the orientations of the loads and internal stresses in the vicinity of the web hole. With the double symmetrical WF shape, the eccentricity to the hole, or the location of the hole w.r.t. the max. normal stresses due to bending is conveniently related to the N.A. which is also the centroid of that section, and they have a well defined relationship to the extreme fiber stress location. In your case the centroid and N.A. may not be (probably aren’t) the same, and this is certainly true with a composite section, in its service state. So, the N.A. is not a very good ref. datum, and d/2 seems meaningless to me at the moment also. I would try to understand the derivation of, AISC’s cookbook recipe (formulas), and then convert this to relate the hole location and size to the location of the max. tensile normal stress; and that should be the bot. of the bot. flg.

Some other considerations: with a nominal sized hole you don’t change the section props. much and you might ignore this change; obviously the closer the hole is to the bot. flg. the higher the normal tensile stresses around it will be, alternatively, if the hole is at the N.A. the primary loss is some shear area capacity, almost no bending stress issue; at a given location the shear forces are constant whatever the hole size or height location; the larger the hole is the greater the normal stress concentration factors are around the hole and now add the highest normal stresses (close to bot. flg.) and you might have trouble. What does AISC say about allowable stress magnitudes around the hole? What do they say about this w.r.t. various hole sizes? They are probably actually using “e” to relate normal stress magnitude around the hole to allowable tensile stress (old guy talk) at the extreme fiber. And, this then ties back to stress concentration factors vs. hole size and a potential fracture stress level which acts at the hole center line in the height direction, and this is a combined stress problem. A rough cut hole adds another set of stress raisers to the problem.

 

[smjohns78]

dheger:

Thanks for the response. You are right, it is always good to understand the origin of the equations in any reference you use. It seems like the equations are based on plastic design. The way they use "e" seems to account for the reduction in the plastic section modulus Z. For non-composite design perhaps "e" should be based on the location of the PNA. However, for composite design, they seem to use the same definition of "e" as for the non-composite.

To respond to some of your comments, DG2 does have limitations on the size and location of the hole. Those limitations generally are in place to ensure that the top and bottom T-sections at the hole have strength/stability. The guide also has requirements for finishing at the corner of openings to minimize effects of stress concentrations.

 

[ishvaaag]

In any case lacking guide support a 3D FEM study may be easy and worth, since just some extrusions, boolean adds and substract, and you may gain clear vision of the stresses. You can compare then with symmetrical cases for the steel only and composite sections and see if the range of variation is much for your typical case. I think it shouldn't be much since the affection is for the web mainly and in any case for the hypotheses that represent the demand both in symmetrical double tees and singly symmetrical you need to have the bending strength -for longitudinal stresses- in place.