T.2.1 Column Calculations: The Columns Are Modelled As Either
T.2.1 Column Calculations: The Columns Are Modelled As Either
1. in its most basic form a beam, within the depth of the slab, with a width equal to the transverse column
width, c1, and a depth equal to the depth of the slab.
2. when a column capital is present then the width is that of the capital.
3. when a drop panel is present then the depth is that of the drop panel and the width of the column or capital.
4. when a transverse beam frames into the joint then the transverse torsional member consists of the beam
projection above and below the slab plus flanges extending on either side of the beam a distance equal to the
projection of the beam above or below the slab but not greater than four times the thickness of the slab.
Torsional Constant
Calculate the cross-sectional torsion constant, C, used to approximate the St.Venant torsion constant as given by
equation T.2.1.
Where x and y are the smaller and larger dimensions of the attached torsional member respectively. The cross-section
is divided into separate rectangular parts and the C of each component part is then added to give the total torsional
constant.
Figure T.2.1
Figure T.2.2
Note:
2. the transverse support width dimension, c2, is generally the column dimension however
where a column capital exists it is taken as the width of the capital at the soffit of the slab.
• slab depth
• average parallel beam depth
• average drop panel depth
The inertia of the column capital is taken as that at mid-depth of the capital projection below the deepest member.
The projection of the capital is assumed to be uniform around the perimeter of the column. The same capital is
assumed to be present at the slab / column connection above and below the slab being analysed.
The stiffness is calculated using moment-area principles. Note that the horizontal centre-line of the equivalent frame is
modelled as the centroid of the slab / beam at midspan, denoted as the datum.
The procedure is summarised as follows (ref 20):
X is the distance to the centroid of the 1/I diagram from the datum. This is obtained by taking first mom
column about the datum and dividing by P.
R is the second moment of area of the 1/I diagram about X
= areas about X
If the column were to have the remote end pin-ended then the stiffness is taken as being three-quarters that given by
the above equation.
The same procedure is followed for the column above and below the slab thereby giving the column stiffness kca and
kcb respectively.
As a check on the above the approximation proposed by Cross and Morgan (ref 19) may be used as described below.
where
Icol = the column inertia remote from the column ends
Ec = Youngs modulus of the column
L = floor to floor height
L' = clear column height
Figure T.2.3
Alternatively Rice and Hoffman (ref 20) proposed
where
h = the depth of the increased stiffness region at the top of the column.
Equivalent Column Stiffness
Combine the column stiffness (kca + kcb = Σkc) and the stiffness of the transverse torsional member (k t) to give the
"equivalent column" stiffness using equation T.2.5.
(T.2.5)
Note that at the edge of the frame the transverse member must be designed to resist, in torsion, a portion of the
moment attracted to the equivalent edge column. RAPT, in its output, distributes the total moment in the equivalent
column to the columns above and below.
If the design is that of a typical strip of a one-way slab then the inertias of the frame are based on the width of the
strip being considered. This then requires that the column stiffness should be reduced to reflect the ratio of the width
of the strip being considered to the full panel width , Lt, at each column. RAPT uses the ratio of the one-way width to
the L2 dimension at each column line. For a one way beam the full equivalent column stiffness's are used.
The column stiffness above and below can be represented in the following simplified form
kc = x Ec Ic / L
where x is a factor defining the stiffness. For a column of uniform cross-section fixed at both ends x would have a
value of 4. Define xa and xb as the stiffness factors for the columns above and below the joint respectively.
Express the stiffness of the columns as a ratio, αa and αb, of the total column stiffness
αa = kca / (kca + kcb)
αb = kcb / (kca + kcb)
If the assumption is made that αa and αb will also be the ratio of the equivalent column stiffness above and below, with
respect to the total equivalent column stiffness, then the equivalent column stiffness's above and below are
keca = αa kec
kecb = αb kec
This implies that the contribution of the transverse member is shared to the two columns in the ratio of their own
stiffness's.
The Iec printed to the screen is based on an equivalent prismatic member using a stiffness factor of 4 ie.
As the length of the equivalent column is undefined we can achieve this result from
For checking of the program using a plane frame computer program, the moment of inertia to be entered for the
equivalent column will be given by the above equation. If columns are to be modelled above and below then the actual
lengths, Ha and Hb, should be used and the akec factor applicable.
The cross-sectional area for the frame analysis is taken as the gross sectional area of the main column section.
Columns above are analysed as having vertical roller supports. If the frame is braced then a vertical roller support is
placed at the two extreme ends of the frame.
These calculations are repeated for each column above and below and then applied to the frame analysis.
This calculation is repeated for each column above and below and these inertias are applied to the frame analysis.
Inertias
Inertias are calculated based on first principals.
Tapers
For tapered elements, RAPT divides the taper length into smaller even segments. (equal segments of approx length =
span/15). The inertia and area given for the taper length are calculated as follows
Frame Application
The calculated inertias for each segment within a span are applied to the frame analysis. The frame is modelled as a
straight line (no vertical members beside the columns) with the increased inertias at the steps modelling the steps in
the frame.
T.3 Blank
For post-tensioned structures the default distribution is performed solely on the factor nominated in Input F1 [Input].
As set out in AS1480 and ACI318, a percentage of the panel moment at each column centre-line and at the maximum
span moment location are obtained from the following tables.
where
(T.4.1)
(T.4.2)
Ib is the inertia of the beam defined by the projection of the beam above and
below the slab plus flanges extending on either side of the beam a distance
equal to the projection of the beam above or below the slab but not greater
than four times the thickness of the slab.
= 0 if no parallel beam
Is is the inertia of a width of slab equal to the full width between panel
centre-lines, Lt, excluding that portion of the beam stem extending above
and below the slab.
C is the transverse torsional constant (as per the column calculations)
L1 is the length of a span in the direction in which moments are being
determined, measured centre to centre of supports.
For cantilevers the percentage along the full span is taken as the value at the exterior support. For internal columns
the a1 and bt values are based on the average slab and beam inertias either side of the column.
Column Strip: Negative Moment Factors for an Exterior Support
75 / 25 55 / 45
75 / 25 55 / 45
The remaining proportion of the total panel moment is taken by the two half middle strips comprising the design strip.
RAPT selects the maximum moment point within each span and applies the above positive column strip factor. The
percentage between this point and the two support centre-lines is taken as varying parabolically A plot of a two-way
reinforced slab bending moment diagram for a middle strip demonstrates this effect.
When there is a beam spanning in the design direction and within the column strip it is designed to resist a percentage
of the load taken by the column strip. The percentage reflects the relative stiffness of the beam to that of the column
strip. The factor is the minimum of the following
0.85 or
0.85 α1 Lt/L (T.4.3)
Normal design office practice and the method adopted by RAPT is as follows.
Where a beam spans in the design direction, references to the column strip within RAPT refer to the beam.
The different codes specify different reinforcement and tendon distribution in two way slabs. ie for AS3600 -
Reinforcement and tendon distribution in two-way flat slabs (Clause 9.1.2 AS3600) requires that at
least 25 percent of the total of the design negative moment in the column strip and adjacent half
middle-strips shall be resisted by reinforcement or tendons or both, located in a cross-section of slab
centred on the column and of a width equal to twice the overall depth of the slab or drop panel plus
the width of the column.
The width of the panel is defined in Input F2 [Spans], at either end of each span. Therefore we may have a different
panel width on either side of each column. In normal design situations these two values will be equal however in some
cases the panel being analysed may step sharply (in plan) at the column line which will result in different panel widths
on either side of the column and hence different column strip widths.
The calculations at each column line proceed as follows. (for all code types)
Note:
1. For cantilevers the column strip width is taken as uniform extending from the exterior column. It may be
argued that the shorter the cantilever the smaller should be the column strip width. The codes do not make
allowance for this.
2. the width of the column strip may vary linearly in plan in each span.
T.5.1 Flexure
In a linear-elastic analysis, column supports are modelled as isolated support lines of zero width. The support is in fact
of finite width and hence the column reaction is not a single point reaction but rather a trapezoidal distributed load
over the width of the column. The peak bending moment obtained from the analysis at the column centre-line is
therefore an over-estimate of the maximum bending moment as the distributed reaction forces will have caused a
reduction in this value.
This effect is certainly present at a section centred over the column, the column strip, but must be reduced as we
move further towards the centre of the panel.
It has been common to allow for this moment reduction effect by designing the slab for the moment at "the critical
section". This section is at the same location in both the column and middle strips. In previous Australian and
American codes the location of the critical section has varied from the face of the column to other intermediate points
within the column width. RAPT uses the appropriate critical section for each code.
In calculating the critical section, RAPT will, if specified by the user in Input Screen F9, take account of transverse
beams. The critical section is taken at the projection of a 45° line from the column face to the under side of the slab /
beam. If this 45° line crosses the side of the transverse beam then the critical section is limited to the width of the
transverse beam.
AS3600
Presently AS3600 (reference 7) states that
"The critical section for maximum negative bending moment shall be taken at 0.7 times the span support length, as,
from the centre-line of the support."
1. for beams or
for flat slabs without either drop panels or column capitals,
the distance from the centre line of the support to the face of the support; or
Note: Drop panels on the underside of beams are therefore disregarded in the
determination of as.
2. for flat slabs with drop panels or column capitals or both,
the distance from the centre-line of the support to the intersection with the plane of the
slab soffit of the longest line, inclined at 45 degrees to the centre-line of the support, which
lies within the surfaces of the slab and the support."
ACI318
Clause 8.7.3 states
For beams built integrally with supports, design on the basis of moments at faces of support is permitted.
RAPT assumes the critical moment occurs at the face of the columns / capitals / transverse beams.
Eurocode2
Clause 2.5.3.3.(5) states
Where a beam or slab is cast monolithically into its supports, the critical design moment at the support may be taken
as that at the face of the support.
RAPT assumes the critical moment occurs at the face of the columns / capitals / transverse beams.
SABS 0100
Clause 4.3.2.1.4 states
For continuous beams over supports, the design hogging moment need not to be taken as greater than the moment at
a distance d/2 from the face of the support.
Clause 4.6.5.1.3 states for flat slabs
Negative moments greater than those at a distance hc/2 (where hc = effective diameter of a column or column head
not > 1.5 * the smallest dimension of the column head) from the centre line of the column may be ignored.
T.5.2 Shear
For shear the critical location is given as
Note: In each code the above conditions apply for generally uniform loads. RAPT does not calculate
any enhancements for shear strength close to the supports when points loads are applied near the
supports. Designers will need to check these relevant code conditions by hand if applicable.
In each case RAPT assumes d = 0.8D (ie d=0.8 x member thickness).