structural steel - discontinuity at base of light pole

huangyhg

structural steel - discontinuity at base of light pole
i reviewed the structural design of a light pole from another engineer.
the pole is a 8" standard (8.625" od) pipe with a 6" diameter hole near the base (cable port).  while cable ports are necessary and common, i've never seen such a big chunk of steel taken out of the cross section at the point of maximum moment.
technically, the stresses work.  that is, the demand vs. capacity is about 90% after consideration of the reduced section at the port (but ignoring the stress concentration effect).  however, i believe the design is just plain bad, and need code ammunition to refuse this nonsense.
does aisc or any other structural steel building code penalize stresses due to discontinuities?  certainly tension connections get penalized in aisc (chapter b) with shear lag factors - how about main flexural members?
check out our whitepaper library.
1. stress concentration. i have noticed edges around openings are thickened.
2. lid of opening does contribute some resistance.
in bending.
it may be a failure mechanism. if there are no safety barriers preventing a car hitting the light pole there will be a slip base, so the pole will fail when hit by a car and prevent the car from being wrapping around the post.
i imagine that the reduced section looks like a c.  so i'm curious.  can the tips of the c go into compression?  the section then would tend to buckle.  were you able to take that into account?
i know light poles and cantilever wind towers are different for analysis but also similar. in wind tower design we have large doors in the side of the towers, to analysis this we look at fatigue fairly heavily with rain flow analysis of the wind loads and then fatigue of the steel tower. generally, the walls are thickened as per dfkhan post with some cool welding details. therefore, if you wanted to get them on fatigue i found an interesting article that make help.
when in doubt, just take the next small step.

miecz,
you are correct that the section looks like a c, and the tips go into compression.
the engineer did a good job of checking combined flexure and compression to define the demand stress at the "tips" of the c in the code check.
we all know that stress flows don't make 90 degree turns so well.  aisc doesn't seem to address this (at least not that i can tell).
in my mind, it would be the same as notching the bottom flange of a simply-supported wide flange beam in the middle of the span, then running a stress check and saying that allowable stresses were less than code, and calling it good, all the while ignoring stress concentrations.
mechanical engineers know better than this.  don't structural engineers?
blodgetts books go into this exact situation.
usually you need to stiffen around the opening.
you may find difficult to extract a stress concentration out of a structural code on a standard shape (well, not so standard after cutting it). in any case you may argue this just on technical considerations. a look on the galambos text that has something that i re  
if i take a sheet of paper, curl it into a c shape and then try to bend it so that the tips go into compression, the tips buckle radially from the center of the c.  granted, the unsupported length of the hole is short, but if the plate is thin, it may buckle long before it gets to yield.  did the designer adjust the allowable compressive stress for buckling?
thanks for the input.
csd72,
you are correct.  ports in poles usually have a stiffening ring around the cut opening.  this opening does not. however, even if the opening is stiffened at it's perimeter (such as a round pipe welded in the wall), i think this mainly stiffens the section.  i'm not convinced stiffening rings help with longitudinal stress transfer, or significantly reduces stress concentration.
miecz,
good comment.  yes, the design did take the slender compression element into account.  opening is not that tall, so buckling is not a limit state of the free edge.
curious.  what did they use for an effective length?  
also, i think a stress concentration factor (of at least 2) should apply to a buckling failure mode.  this is just a gut feeling.