huangyhg

cantilever steel beam at column
when you design a cantilever steel beam with moment connections at both sides of the column, do you assume part of the moment goes to the column (due to relative stiffness of the beam and the column) and design the column for that additional moment?? please advise. for some reasons, we have been assuming the beams are having two pinned connections, and designed columns with vertical load only. now, i am worried...

by definition, a cantiliver is an element with one end fixed and the other end free, therefore you have moment in the fixed end, so, the element providing support for the cantiliver it is in this case, your column, and has to be design to both the axial loading and the moment by the cantiliver...
why do you think a beam analyzed as pinned-pinned-free is not valid? if your connection is free to rotate, a moment should not develope. i certainly agree with taking precautions... however, a cantilever is not only bound by one definition - "one end fixed and the other end free."
don't understand why the moments at column ends should depend on the relative stiffness of beam and column, as a cantilever beam is statically determinate and should not influence by its stiffness the load distribution in the rest of the structure: perhaps you should clarify your structural scheme.
anyway in my opinion you are always safe if you calculate moment connections as if they were pinned, provided you are not in a high fatigue environment and you take the same assumption in all the structural checks performed on the element. however this would of course be a waste of effort in building a moment connection that is not used as such...
prex
if it is stub beam projecting from a column it must have a fixed connection (i.e. able to develop a moment) otherwise it will fail. you should not let all the theory get in the way of what actually happens. you should design the column for the additional moment due to the load from the cantilever beam and make sure that the plate, bolts or welds and the local column capacity are good enough. always put stiffeners in so that the flanges of the beam are continued into the web of the column.
prex - you are right that a cantilever is a statically determininate element...but the moment at the end of the cantilever then extends out to whatever will take it. in this case, you have two - the remaining beam and the column. if the connection is fully moment resisting for both the beam and the column, then the cantilever moment will be distributed to each according to their stiffness.
if the cantilever/beam connection to the column is not designed to take moment, then you could assume that all the cantilever moment extends into the beam.
however, very few connections of any kind are truly "pinned" so i would include some column moment in my column design.
i concur with jae. in such situations, we must consider the difference between what is on paper and what actually occurs.
i might add, however, that the answer to this question relys primarily upon the connection of the column to the beam. does the connection allow internal distortion, thus preventing the column from participating in the moment distribution?
jae
i agree with your first statement. if the inner beam is moment-connected with the column, the moment from the cantilever will be shared by the column and the inner beam.
but i am afraid your second statement is not correct. if the column to inner beam connection is a shear connection, the moment will be totally resisted by the column only. the logic is as below.
first of all the cantilever beam has to be moment-connected with the column, as otherwise the cantilever will become unstable. hence the moment will first get transferred to the column at the joint. if the adjacent beam is not having a moment connection with the column, the inner beam end will tend to rotate and only the bending stiffness of the column will contribute to the resistance.
the inner beam will get the moment only if either the beam has a moment connection with the column or if the cantilever beam is continuous over the column and connected to the column at its bottom flange through a bearing plate. in the latter case, the column will act like a knife-edge without taking any moment from the beam.
trilinga - i agree with all that you say. my second statement, however, was not directed at a shear connection between the main beam and the column, but rather a connection where the whole beam extends over the column (the column perhaps has a cap plate that is simply bolted to the bottom flange of the cantilevered beam. in that case, the moment will most certainly extend directly through beam, past the column.
but a bolted cap plate to a bottom flange has some capacity to accept moment from the cantilevered beam above...just difficult to determine.
you are right jae. that is precisely the arrangement i have mentioned in the last sentence of my post. but, it is not posible when the column extends above the beam level. in such case the adjacent beam has to be either moment-connected or shear-connected to the column resulting in the corresponding moment transfer mechanism.
thanks everyone for your input. usually if the cantilever is short or the beam is a lot stiffer than the column. the moment distributed to the column is small, thus it's safe to ignore. however, in my case, i have a 10'-0" cantilever and the column is a w14x500. i have moment connection (bolted or welded) at the cantilever and the back span. the mega column tends to suck up a huge amount of moment.. which was not anticipated at the beginning. anyway, i have run a computer analysis on the system and re-designed my column with the additional moment. fortunately, it has extra capacity for that. phew. thanks everybody. =) have a nice day.

__________________
借用达朗贝尔的名言：前进吧，你会得到信心!
[url="http://www.dimcax.com"]几何尺寸与公差标准[/url]