Combined actions - Bending and Shear, Bending and Compression, etc

[CELinOttawa]

Morning All,

When do you apply combined actions? The classic case of a simply supported beam is what most of us were first trained on, but this results in a (forgive me) simplified understanding of the stresses in the beam. Further to this, most engineers' "first" beams are either Steel or Concrete; Materials which lend themselves to a high tolerance of "forgetting about" combined actions. Even more, most of us design to cases where we have simplified UDL loadings, again helping to simplify the problem and potentially lead to issues...

Every code I can think of requires a check for combined actions, but when I have reviewed other engineers' calculations, it is rare that I find the check completed, even when I feel this is obviously needed. My gut instinct tells me it isn't always needed, but is that not at the designer's risk and technically a code violation?

For clarity regarding my position, I'll start the discussion by pointing out that I apply combined actions checks in the following situations:

- Point loaded beams
- Beams continuous accross a support
- Columns under combined loading (so most, other than incidental or mandated minimum eccentricity)
- Cold Formed Steel design (all work other than simply supported beams)

Oh, and just to spice up the conversation, I have been the forensic engineer on two investigations where a failure to apply combined actions resulted in collapse. One in CFS and another in timber with a very significant cantilever accross a masonry support wall below.

 

[AELLC]

1) That is easy, I input all point loads on my Excel analyzers.

2) That is a matter of using Beamanal or similar excel spreadsheet to analyze, just a little more work than simple support.

3) Struct. steel and concrete columns that are part of moment frames, etc., you need at least RISA-2D to analyze those. Wood columns supporting 2 beams of vastly different reactions - I use 2 columns adjacent to each other, such as (2)-2x6 next to a 6x6 post. Minor difference of reactions - I use accidental ecc= d/6

Steel columns not subject to moment - I have a minimum eccentricity programmed into my Excel program, I just don't recall offhand what it is.

4) Cold formed - complex problems - haven't messed with those. Too many weird buckling modes. If I were to analyze one of those, I would buy a software that specifically analyses and designs light-gage cold-formed.

So, I think the implication is: If you don't have computer-automated methods, you tend to dangerously simplify designs in the rush to get projects finished.

The definition of a structural engineer: overdesign by a factor of 1.999, instead of the usual 2.

 

[Lion06]

I typically only look at combined loading for axial and bending. There may be a rare case out there where combing shear and bending or axial is appropriate, but I haven't done it yet. The max shear stress is typically at the centroid of the section and the max bending stress is typically minimum at the same location, so their interaction is minimal in my mind (for most scenarios).

 

[CELinOttawa]

I've been giving this some thought, and I still think there is cause for more application of combined actions than I see most engineers undertaking.

AELLC: Thanks for the thoughts; I think you're spot on. That's exactly how most engineers deal with this issue - Let the computer worry about it. BUT, isn't that a little creepy and worrysome about the future of our profession? I've always been a big opponent of the "black box" effect of having a computer do work that someone doesn't know or understand...

Lion06: Your breakdown of the issue is perfect, for Elastic design. If you are basing your work on plastic design, the argument that shear stresses are in the web (or near the centroid) and flexural stresses are at the extreme fibre (in the flanges, or the extreme of the section) is moot. Fully plastic stresses are throughout the section, and the shears then add to the problem.

I am the first to admit that situations where combined stress analysis is necessary are rare, but i have seen them. I would love to see a good paper or chapter or guideline clearly discussing this issue. In the meantime, I'm going to keep applying combined actions whenever I have a significant source of two actions in the same location using plastic design.

 

[ajh1]

AISC does not actually mandate bending/shear interaction checks. The place these become critical is on continuous framing, particularly where change of cross-sections occur. The AISI spec does require a bending/shear interaction check. This check controls most frequently at the end of lap condition where members are framed continuously over an intermediate support such that the moment of inertia is roughly twice as large over the support in the lapped area as it is just outside the lap area. Cantilever members are another area where bending/shear would play a large roll in the effect on the member.

 

[CELinOttawa]

I'm surprised to hear that AISC doesn't mandate bending & shear check when I know CISC requires it for situations where there is high demand for both bending and shear, as well as NZS 3404 has the same requirement. For an example of a situation where this type of check is required, one need only think of a transfer beam carrying multiple floors to omit only the lowest column.