Stiffness Modifiers

General
A number of you tube channels and lectures have discussed this topic. Here, I wish to go from basics.

As far as the IS codes are concern, this was introduced first in IS:1893-2016, i.e., in earthquake design
code for structural analysis. It stipulates:

For structural analysis, the moment of inertia shall be taken as:

(a) In RC and masonry structures: 70% of Igross of columns, and 35% of Igross of beams
(b) In steel structures: Igross of both beams and columns.

Thus, for RC structures, for columns, I = 0.7 Igross and for beams I = 0.35 Igross. And for steel structure no
reduction is specified.

Here, 0.7 and 0.35 are called stiffness modifiers or property modifiers for moment of inertia.

Why is concrete degraded and why not the steel?

Steel is a perfect homogeneous and isotropic material. Its modulus of elasticity is same and constant
for tension and compression. It is a single material and does not crack while loading. Thus, steel section
is keeping its moment of inertia constant for loading up to yield loads. Extremely small deformations
are not taken into account in these considerations.

Concrete on the other hand possess a number of extremely small cracks right from its manufacture.
Under the compressive loads, these cracks will tend to close and hence, the concrete subjected to
compression behaves like a homogeneous material. When it is subjected to tension, due to cracks, its
tension carrying capacity is drastically reduced because the tension can be transmitted only through
uncracked concrete. With application of tension on concrete, the small cracks will open and will reduce
the uncracked concrete matrix. This will further reduce its tensile strength. What we get is the tensile
strength of concrete as hardly one-tenth of its compressive strength.

We conclude: Under tensile loads, the uncracked matrix of concrete reduces and hence for a given
section, the moment of inertia also reduces.

Effect of loading on RCC sections:

Axially loaded members:

The reduction in uncracked concrete matrix is very high. As a result, the tensile strength of concrete
reduces drastically. For comparison, a rectangular concrete section with M20 grade concrete has
characteristic compressive strength of 20 N/mm2, while its characteristic tensile strength is 0.4sqrt(fck)
= 1.79 N/mm2.

Members loaded in flexure:

While calculating strength of the member, i.e., moment of resistance of the member it is customary to
neglect the tensile strength of the concrete. In derivation of flexure formulae, we consider an
assumption that concrete is assumed as cracked in tension (IS:456). However, when we calculate the
surface crack width of the member, we consider a small resistance of concrete in tension (refer to crack
width calculation formula (annex F of IS:456).
Loading the beams and slabs (Flexural members):
Consider a rectangular simply supported beam subjected to UDL for the discussion. With zero load on
beam, not even the self-weight, this RC beam contains plenty of cracks but with limited width and
concrete is called uncracked concrete. With removal of formwork, loading is commenced and beam is
subjected to flexure. The bending moment being zero at supports and maximum at the centre. For
very small loads (i.e., moments), the beam behaves as a homogeneous beam. Because of the presence
of the steel reinforcement, the neutral axis is situated at a distance of more than 0.5 d from the
compression face. If fct is the tensile strength of concrete, and Z is the section modulus of the beam,
the beam can resist a moment of M = fct × Z, with limited cracks. This discussion is not exact but for
understanding the behaviour. When the applied moment becomes more than this moment, as per our
assumption the concrete cracks in tension and not capable of resisting any tension. This moment is
called cracking moment Mcr, of the beam. Note that this happens at the point of maximum bending
moment i.e., at centre of this simply supported beam. Now all of a sudden, the whole tensile force
carried by concrete is transferred to steel. If the steel is not capable of taking this tension, it will snap
and the beam will fail. Thus, Mcr determines the minimum reinforcement required by the beam.

With the increase of the loads (consider gradual increase), the moments at all sections along the length
of the beam will increase (except supports in this case). As per analysis, the compression in
compression zone and tension in tension zone will increase. The depth of neural axis will reduce to
increase the lever arm. Let the UDL now reaches to ws, the service loads for which we are designing
the beam. The bending moment at centre is 0.125 ws l2 and varies parabolically to zero at the supports.

It is very important to understand here, that the depth of neutral axis is minimum at centre
and maximum at support. Since moment is zero at support, the depth of neutral axis will not change
at the section at support. This means:

1. Moment of inertia of section for this simple beam is minimum at centre and maximum at the
support.
2. In general, moment of inertia of the section varies along the span and is minimum at the point
of maximum bending moments.
3. The beam no longer remains prismatic (constant section and constant moment of inertia
throughout the length) and becomes non-prismatic (a beam with variable section and variable
moment of inertia).

Loading the columns and walls (Flexural members):

Columns and walls are basically compression elements. They are primarily subjected to axial
compression and also some moments. When only axial compression is acting, the neutral axis is at
infinity. When moment is also acting on the section, the neutral axis shifts towards the section. With
increase in moments, it will enter the section and goes towards the compression face. The moments
are maximum at beam column joints and zero near the centre of the column length. Similar to beams,
the variable moments will reduce the moment of inertia of the section that is varying along the length.
However, the reduction in moment of inertia is not as appreciable as for the beams. This is because
always some axial compression is acting that has a tendency to close the cracks. This means:

1. Moment of inertia of sections varies along the span and is minimum at the point of maximum
bending moments.
2. The column no longer remains prismatic (constant section throughout the length) and
becomes non-prismatic (a column with variable section).
 3. The reduction of moment of inertia in sections of columns is reasonably smaller in columns as
compared to those in beams.

Difficulty in analysis:
The analysis of the structure by using methods such as moment distribution, matrix methods, or finite
element methods cannot be done by routine methods of assuming prismatic member. And corrections
are necessary.

To overcome this difficulty an approximate method of considering equivalent moment of inertia that
is constant for the whole length of the member is adopted. This means that the equivalent moment
of inertia of a loaded member is one which is calculated by using some multiplying factor to the
moment of inertia Igross of the unloaded member. We can write

A non-prismatic beam can be replaced by a prismatic member by applying a stiffness modifier to the
gross moment of inertia.

Thus, Stiffness modifier = Iequivalent of the member / Igross of the member.

It is always less than unity.

Understanding the concept of factored loads:

In design of any structure using elastic theory or working stress method, we used permissible stresses
in the materials. The stress in the material is not exceeded to the permissible stress. These stresses (in
bending, shear, tension, etc.,) are derived by using some factor of safety on

1. the yield stress or proof stress when material is ductile, e.g., steel bars, steel sections, etc.
2. the failure stress when material is brittle, e.g., brick, concrete, wood, etc.

The factors of safety in second case are much higher than the first case.

Next, we learnt that the loads (or deformations) applied on the structure or a member of the structure
is a prime cause of the failure. And we attended the non-linearity of stress-strain diagrams of materials
used. We concluded to use loads for design. Thus, a beam is observed to the failure by increasing the
loads and draw a load-deformation diagram for the purpose. We then devised a method where we
assume that the service (or characteristic) load will increase to the extent the acceptable capacity of
the structure can be equated. This load, we call as a factored load. This is not a real load, but a
conceptual load calculated by applying a partial safety factor on loads. We apply this load and equate
the capacity of beam to sustain this load. Thus, the factored loads are arrived at by multiplying the
corresponding service load by its partial safety factor of loads.

Strength design and Serviceability design:

In strength design, it is considered that member should be strong enough to carry the applied
loads.

In working stress method, the loads are service loads. In limit state method, the loads are
factored loads, where the partial safety factor is always more than unity. The examples are:

1. The net bearing pressure on soil shall not exceed the allowable bearing pressure on soil (WSD)
2. The maximum shear stress on a circular shaft shall not exceed the permissible stress on the
shaft material (WSD)
3. The member should be strong enough to carry the applied factored loads (LSM).
 These examples give understanding of strength design. For RC designs of structures, case (3)
is applicable.

In serviceability design, we check that the member should not deform (e.g., settlement,
rotation, deflection or cracking) excessively so that its function is not served.

In working stress method, the loads are service loads. In limit state method, the loads are
factored loads, where the partial safety factor is equal to or less than unity.

The examples are:

1. The settlement of soil shall not exceed the allowable settlement of soil (WSD)
2. The angle of twist of a circular shaft shall not exceed the permissible twist of the shaft (WSD)
3. The member should be stiff enough to limit the applied deformations (deflection and cracking
of the member) at the applied service loads (LSM).

These examples give understanding of serviceability design. For RC designs of structures, case
(3) is applicable.

Variation in stiffness modifiers:

As we have seen that stiffness (moment of inertia) continuously reduces with loads.

This means:

SM at ultimate loads < SM at service loads < SM at 0.5×service loads … … and so on.

Design of RC members is carried out for (i) service loads and (ii) strength loads.

Therefore, we have two different stiffness modifiers as follows:

Stiffness Modifiers to be used in analysis as per IS: 16700

Element Serviceability design Strength design

Slabs 0.35 0.25

Beams 0.70 0.35

Columns 0.90 0.70

Walls 0.90 0.70

Here Serviceability = Working loads

Strength = Ultimate loads.

The section modifiers for cross section area in all cases are 1.0 This means that no reduction in cross –
section areas to be considered.

The factors given above shall be used in all directions, viz., Ix, Iy, and Iz (torsion). Some codes of practice
give little less values for torsion modifier.
Misconceptions, reasons and corrections:
With introduction of these modifiers, many misconceptions prevail. Some of them are discussed here.

1. Stiffness modifiers shall be used with load combinations involving seismic loads.

Since these modifiers were introduced in seismic code for the first time (IS:1893), they shall
be used with seismic loads.

Correction

Stiffness modifiers shall be used for all primary and combinations of loads.

2. Stiffness modifiers are used with unfactored loads and factored loads.

This is because these were introduced in IS:16700-2017 as with unfactored loads and factored
loads. Later in IS:16700-2023, this is corrected as serviceability and strength.

Correction
Stiffness modifiers shall be used in analysis for serviceability and for strength.

3. Stiffness modifiers shall be used in analysis as well as in design.

Correction
Stiffness modifiers shall be used in analysis for serviceability and for strength and not in design
of the member.
Reason
In design of the RC member, right from its beginning, we have assumed no tensile strength of
concrete and prepared design formulae. There is no meaning in using these modifiers in
design.

4. Stiffness modifiers lead to uneconomical design.

Comparison of latest adopted design by code and earlier adopted design by code itself is
wrong. We need to accept and respect the latest changes in codes of practice that are the
outcome of experienced and learned people.

5. Stiffness modifier in torsion shall be assumed very small to get no or small torsion.

Correction

Stiffness modifier for torsion shall be the same as that for bending as per code.

Reason

Many learned people use smaller modifier to reduce the torsion and get economical design.
Since, code has not clarified we need to use the same. Also note that concrete does not behave
as per our design, rather we design it as per its behaviour.

Determinate structural elements:

Stiffness modifiers are not necessary to be specified for determinate structures like simply
supported beams. This is because there will not be any moment distribution in such elements.

The structural elements of foundation like continuous foundation beams are determinate
structures. These beams are supported by columns and walls that are supporting the building.
 The reactions of the columns on continuous foundation beam are nothing but applied loads.
Thus, foundation beams are determinate and do not require the modifiers to be specified.

Note that if you specify the modifiers to determinate structures then also there will be no
change in results since there is no moment distribution required at joints.

Understanding of analysis for strength without and with stiffness modifiers:

I/L = 14 I/L = 14 I/L = 14×0.35 I/L = 14×0.35

I/L= I/L= I/L=

I/L=8 I/L=8 I/L=8 8×0.7 8×0.7 8×0.7

0.64 0.39 0.39 0.64 0.47 0.32 0.32 0.47

-100 100 -100 100 -100 100 -100 100
64 0 0 -64 47 0 0 -47
0 32 -32 0 0 24 -24 0
-
-36 132 132 36 -53 124 -124 53
0.36 36 0 0 -36 0.36 0.53 53 0 0 -53 0.53

(a) Stiffness modifiers not used (b) Stiffness modifiers are used

Observations:
At beam column joints, the bending moment increases at ultimate loads that are used for
strength design. The moments at the centre of beam will correspondingly be reduced. Note
that this is the correct behaviour of an RCC frame.

The fundamentals behind section modifiers for slabs and walls are also can be described with
similar manner.