Need a book? Engineering books recommendations...

Return to index: [Subject] [Thread] [Date] [Author]

Re: Column design

[Subject Prev][Subject Next][Thread Prev][Thread Next]

Though I do not agree with the concept of designing the column for no moment (this happens a lot in SE Asia as well), you will find that the end columns on the first floor above ground for every low-rise concrete framed building you have ever designed will have theoretically formed plastic hinges. When you have designed the building for gravity load effects, the moments calculated are relatively small compared to the shrinkage and temperature shortening moments. If you were to add these moments into your design then all of your end columns will "fail" and form plastic hinges.

As long as the building is braced by other mean, normally shear walls, this is not really a problem, as the columns will still provide vertical force capacity and the shear walls will provide sway capacity and the floors will act as diaphragms between.

The advantage our zero column moment designers have is that their floors are designed assuming the column is attracting no moment and are therefore conservative in terms of strength and deflections (if they have actually calculated any). From my experience these designers still provide some top reinforcement at the top of the beams/slab at the end columns so they do have crack control there. If the column bending stiffness is, for some reason, reduced or disappears completely, the floors are ok. Conversely, designers who design the floor assuming end column moments are designing floors assuming that column stiffness will be there and will therefore be unconservative in the floor design if that stiffness were not available for some reason.

Then you get someone designing a column under a transfer beam and ignores column bending stiffness in the calculations when the column is trying to attract about 15,000KNM of moment. Then you have problems.

At 03:27 PM 21/12/2005, you wrote:
And I will point out that they are not then really designing "per" ACI
318...which is fine.  However, the person who posed this question said
that the region used the British code and ACI 318 as their code (aka law
in the region) or at least as the minimum guideline.  If done "per" ACI
318, it is rather difficult to detail a monolithic beam-to-column concrete
joint such that there will NOT be moment transfer to the column.  And it
ain't cause I am trying to "be smarter" than them cause I ain't smarter
than the laws of physics/nature...if the joint is detailed such that it
will attract moment to the column, then it don't matter how you, I, or
anyone else models it in some fancy program...the column will still
attract moment.

And I don't disagree with the notion that reinforcing steel could yield
and a plastic hinge will form at places.  I understand the concept
perfectly well.  The problem is that _IF_ there is steel there to yield
(i.e. assume we are talking about negative beam moment reinforcement at a
monolithic beam-to-column joint), then there will be a moment in the beam
that will max out at the "plastic moment" at the location.  And that
moment does not just magically go has to go SOMEWHERE.  And if
we are talking about an exterior column, then the only place that moment
can go is the column.  And again, it ain't cause I am smarter than anyone
else.  So, in such a situation, if the column is not designed for gravity
load moments and the actual moment resistance of the column (whether you
or I design a concrete column for a moment it will have moment resistance
even if it is for some crazy reason unreinforced but especially if it has
longitudinal steel [which would likely be there for axial load at a
minimum]) if such that it is smaller than the "plastic moment" of the
beams negative reinforcement, then more than likely the plain simple fact
(again irrelevent of what you or I design or want) is that there will be a
plastic hinge that forms in the column before one forms in the beam.  And
conventional wisdom in structural engineering is that typically hinges
(plastic or otherwise) in columns usually ain't a good thing.

So, the only typical way to get a column in concrete contruction to not
take moment from negative bending (at least that I can think of the top of
my head) from a beam is to not use monolithic concrete construction...that
is go to precast.  In which case, you are now talking about haunches or
corbels...and in which case you would STILL need to design the column for
moment as now you will have eccentric loads applied to the column.  Now,
in theory, you could do a monolithic concrete frame and just not have any
beam negative reinforcement at the beam to column joints...but this is
generally not a preferred style of construction.  At a minimum, you will
now likely have deep flexural cracks at top of all your beams where they
hit the columns...and depending on this could affect the shear capacity at
the end of the beam (i.e. if the load is enough to open the crack wide
enough for enough of the depth of the beam, then you may not even have
enough shear friction [i.e. due to lack of aggregate interlock due to the
open crack if wide enough] to resist the shear at the end of the beam).
So, if there is ANY negative moment steel (where it is designed to be
negative moment steel or just steel added for some other reason such as to
ease placing of stirrups or to provide steel for shear friction/aggregate
interlock so that there will be shear friction capacity at the top of hte
beam too), then there will be negative moment in the beam, and where there
is negative moment in the beam, there more than likely there will be
moment in the column, especially if it is an exterior column.

As to the 2x4 comment, I am not sure how similar the situation is.  In the
cases of 2x4 rafters, you even point out that there is a possible
"secondary" mechanism that could be being "neglected" that could explain
why there has not be many notiable problems.  In general, the
beam-to-column monolithic concrete connection is a much better "known"
animal and as such has fewer things that we "neglect" that could be
significant "secondary" mechanisms.  In addition, there is always the
issue of how many of those 2x4 rafter systems (or non-moment designed
column that you say are prevalent in the mid-East area) have been exposed
to significant or code level loads.  Heck, I can design my roof out of
cardboard if I want and it will work...up to a certain level/load.  If I
go for 10 years (or more) and the load never comes for some miracle, it
does not means that it was designed well...just means that I have been
lucky that a large enough load has not come along to destroy my nice
cardboard roof structure.  And last, even though some people cannot
explain how or why those 2x4 roof rafter systems have worked in the past,
how many of those people would design a roof system for a building today
with 2x4 rafters (assuming we are not talking about a rather short span
where 2x4s might actually work)?  The point is that even though there is
some historical "proof" that they work for some reason that cannot be
explained too well, conventional wisdom today says that it ain't a good
idea so I am betting that you would be hard pressed to find an engineer
who designs 2x4 rafters for today's buildings.

As to this situation, hey, those in the Middle East want to use the "same
short cut calculations" and have them be accepted, then that is certainly
their right.  And it is my right to believe that it ain't such a good idea
and to point out that they are NOT likely then designing per ACI 318.
Now, I certainly don't know all the details of the particular situation
and as such there could be information on this specific project (and
others similar to it) that I am not aware of that could justify the notion
of not needing to design an exterior column for moments.  But, based upon
the limited information that has been provided, this does NOT appear to be
the case.  And as such, it sounds as if I were in this guy's place, I
would be designing the columns for moments no matter what the typical
practice in the region is.  Maybe that would mean that I would not have
much work as people would go to those who don't design or require that and
thus end up with designs that cost less to build than what I might design.
If so, then so be it.  Just means that I would either have to find work
else where in the world where I would not be perceived as such a
conservative/costly engineer or find another profession.  But, I am not
gonna design something that my education and experience tells has the
potential for problems unless someone is fully able to convince me that I
am wrong in my current knowledge or belief.


Adrian, MI

On Tue, 20 Dec 2005 ASQENGG2(--nospam--at) wrote:

> The issue is not the eccentricity of loadings, the column has to be designed > with the moment due to eccentricity no matter what. The issue is the ability
> of the column to attract moment from the beam base on the gross  sections
> between the beam and column.
> The fact of the matter is most of the Middle East use the same short cut
> calculations and is being accepted. It is similar to accepting conventional
> framing in wood construction even though if you follow the regular code for wood > most of shear wall won't pass. The acceptance such practice is the fact that > when the steel yield and develop a hinge in the joint then the beam is still
> capable of supporting the vertical load being design for  M=wL^2/8.  And so
> the practice has been done for many many years and if a new engineer trying to
> be smarter than the rest of the region I think  it is being bit too cocky.
> When in Rome do what the Romans do.
> Another example are the old houses that use 2x4 rafters.  If you  calculate
> base on the present code, then the 2x4s won't pass. Some engineers think that
> the rafters held up for many years due somewhat effect of  roof sheathing
> acting as shell.  Some attributed it to some degree of  redunduncy.
> In a message dated 12/19/05 4:39:44 P.M. Pacific Standard Time,
> smaxwell(--nospam--at) writes:
> Yeah...but for a column (end/exterior column or interior)  "just
> supporting a pre-cast beam" will still have the load from the beam  applied
> eccentrically to the column (unless if is a roof  beam-to-column
> connection, in which case it could be a concentric  load).  And that
> eccentric load will result in the moment in the  column (not matter what
> "load pattern" used for an exterior column but only  for "unbalanced" loads
> for an interior column).
> You could in theory  have the beam (exterior) or beams (interior) sit on
> the column and then  have the next stories column sit on the top of the
> beams.  But you  would still have some eccentric load cases for the
> interior columns.   And detailing such a beast so that the beams can just
> sit in bearing, but  the column can not translate laterally relative to
> beam, is rather tough  thing to do.
> The real point is that typical reinforced cast in place  construction you
> design the concrete system as a frame.  This includes  exterior columns.
> This because typical CIP concrete construction has all  the beam to column
> joints as continual, integral concrete...and thus, such  joints cannot
> really behave anywhere close to a pinned  connection.
> Regards,
> Scott
> Adrian,  MI

******* ****** ******* ******** ******* ******* ******* ***
*   Read list FAQ at:
*   This email was sent to you via Structural Engineers
*   Association of Southern California (SEAOSC) server. To
*   subscribe (no fee) or UnSubscribe, please go to:
*   Questions to seaint-ad(--nospam--at) Remember, any email you
*   send to the list is public domain and may be re-posted
*   without your permission. Make sure you visit our web
*   site at:
******* ****** ****** ****** ******* ****** ****** ********

Regards  Gil Brock
Prestressed Concrete Design Consultants Pty. Ltd. (ABN 84 003 163 586)
5 Cameron Street Beenleigh Qld 4207 Australia
Ph +61 7 3807 8022              Fax +61 7 3807 8422
email:          gil(--nospam--at)
email:          sales(--nospam--at)
email:          support(--nospam--at)

******* ****** ******* ******** ******* ******* ******* ***
*   Read list FAQ at:
* * This email was sent to you via Structural Engineers * Association of Southern California (SEAOSC) server. To * subscribe (no fee) or UnSubscribe, please go to:
* Questions to seaint-ad(--nospam--at) Remember, any email you * send to the list is public domain and may be re-posted * without your permission. Make sure you visit our web * site at: ******* ****** ****** ****** ******* ****** ****** ********