Need a book? Engineering books recommendations...

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

Fw: Column Splice Locations

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

Good catch!

By my interpretation, it looks like you are correct in that UBC97 DOES
require splices to be placed within the middle half of the column if

1)  (Ag*f'c/10) < Pu < 0.3Po
2)  Mu<phi*Mn

But strictly speaking, it does NOT require the same if

3) Pu<(Ag*f'c/1)  OR  Pu>0.3Po


4) Mu>phi*Mn

Since section 1921.4.3 is not referenced for these cases.

First part of 3) makes some sense, in that the frame member is not
considered a column (according to the UBC) unless Pu>(Ag*f'c/10)  (see
section 1921.3.1.1 and section 1921.4.1).  So, if Pu<(Ag*f'c/10) for ALL
load cases, it isn't a *column*, so the column provisions don't apply (I
personally would be careful with this way of looking at a member, but I
guess the UBC is relying on the engineer to intepret these provisions, and
all provisions, in light of the structure being designed and the engineer's

But it doesn't make much sense, IMO, for the UBC not to require the same for
the second part of 3), and even more so for 4).  I would have to believe
that this was an unintended omission, esp for 4), since a column
(Pu>(Ag*f'c/1)) which will likely undergo inelastic deformation (even if it
is less than foreseen for a fully ductile SMRF) SHOULD have the splice in
the center.  Take for example a hypothetical structure which has a perimeter
SMRF and interior frames which *don't resist seismic loading*.  If the
interior columns, when checked for maginified deformations, show that they
will yield, how are they different from the perimeter frame columns, and why
should they be designed differently?

I would say that UBC97 was probably trying to create a safer situation by
requiring middle located splices for columns which are likely to experience
some hinging, in order to keep the splices from having to endure plastic
deformation (which they cannot do very well, even if they have closely
spaced hoops - depending of course on the level of plastic deformation
required on their part).  But it would seem that these columns were divided
into 3 categories, and only 1 of these was actually covered.  I'm not sure
if perhaps a column with high axial load (Pu>.3Po) having a splice in a
hinge region would perform better than one with low axial load (haven't seen
a studies on that), but maybe it would so the second part of 3) may be
correct.  But I still say that 4) makes no sense.

If I were designing a column which *didn't resist seismic loads* (I do that
because I don't believe it can ever happen.  If a column is in a structure
undergoing an EQ, it WILL have moments/shear/axial load induced in it, to a
varying degree depending on stiffness, because it has to deform along with
the rest of the structure.  The EQ won't care if it has been designated as
"non lateral load resisting" and neither will the structure.  It then
becomes a matter of choosing an appropriate level of deformation and hence
corresponding moments/shears/axial loads, the UBC magnified def approach
being one way), I would probably go ahead and put the splices in the center,
unless the other elements in the structure, such as shear walls, braced
frames, or very stiff frames, were signicantly stiffer than this column.

Anyone else?

T. Eric Gillham PE
GK2 Inc.
PO Box 3207  Agana, Guam  96932
Email - gk2(--nospam--at)
Ph:  (671) 477-9224
Fax: (671) 477-3456

-----Original Message-----
From: Gregory P. Luth <gluth(--nospam--at)>
To: SEAOC (E-mail) <seaint(--nospam--at)>
Date: Monday, January 11, 1999 10:50 AM
Subject: Column Splice Locations

>Section 1921.7.2.2 of the 1997 UBC applies to columns that do not resist
>seismic loads, with axial load between 0.1 f'c and 0.3Po, in which the
>compatibility moments due to Rw*delta do not exceed the design moment.
>By reference to 1921.4.3, it appears that this section requires that the
>lap splices in such columns be located in the middle 1/3, a significant
>departure from previous codes.  Is my interpretation correct?  Can
>anyone explain why this provision is not triggered for axial loads less
>than 0.1f'c (1921.7.2.1) and for members with axial loads exceeding
>0.3Po (1921.7.2.3)?