Calculation and Comparison of Effective (Cracked) Stiffness values using AQB and ASE modules
Abstract
Linear analysis in SOFiSTiK uses the gross section stiffness, which allows load cases to be superposed but ignores cracking. Where a more realistic estimate of deformations is needed, the effective (cracked) stiffness must be used instead. This article compares two ways of obtaining it: a section check with non-linear stress-strain curves in the AQB module, and a second-order material non-linear analysis (TH2 with NSTR) in the ASE module. Both are applied to a 20 m M60 cantilever column with a 1500 × 1500 mm reinforced section under axial load and biaxial bending.
Introduction
The linear analysis of the model considers the gross stiffness (EI_gross) of the section. This allows the superposition of the results for various load cases through simple summation. While this method is fast and simple, there are cases where the effective stiffness is required instead, for a more realistic and economical design. The effective stiffness (EI_eff) uses the cracked moment of inertia where the extent of cracking depends on the actual force values. Hence, the following prerequisites must be met to evaluate the cracked moment of inertia:
- Cross section should be defined with reinforcements
- Governing values of the internal forces i.e., axial force and bending moments must be calculated through linear analysis.
- Using governing force values, non-linear analysis must be performed for the cross-section.
Two procedures can be used for determining the effective stiffness values in SOFiSTiK. First procedure involves using AQB module in which section check is performed for governing internal forces by incorporating non-linear stress-strain curves through NSTR command. Second procedure involves using ASE module in which non-linear analysis is performed for the relevant load case number.
In this article both methods are applied for a simple case of cantilever column, and the results of effective stiffness values are compared.
Model Description
A cantilever column of 20 m height is modelled with M60 grade concrete i.e. Young’s modulus E = 36887 MPa (Figure 1).

The column has a rectangular section of 1500 mm × 1500 mm. The reinforcement of 32 mm dia. of Fe550 is provided at spacing 100 mm and effective cover of 51mm (35mm+32/2mm) (Figure 2).

Loading
The point loads at the free end (Load case no. 1) are applied resulting in axial force and bending moments at the fixed end. The forces are as follows:
- P1 = 250 kN – Resulting bending moment MY = 5000 kNm
- P2 = 500 kN – Resulting bending moment MZ = 10000 kNm
- P3 = 5000 kN – Resulting axial force N = 5000 kN (compressive)
SOFiSTiK Command
+PROG SOFILOAD
HEAD Loading
LC 1 TYPE NONE
NODE 2 TYPE PP P1 250 P2 500 P3 -5000
$Point loads at top of column producing following forces at the fixed end
$Axial Force = 5000kN (Compressive)
$Bending Moment MY = 5000kNm
$Bending Moment MZ = 10000kNm
endCracked Stiffness
Using AQB module the section check (for cross section 1) is performed for the resulting internal forces at fixed end i.e., N = 5000 kN (comp.), MY = 5000 kNm, MZ = 10000 kNm.
In the report the values of Ey-eff and Ez-eff are shown under “Non-linear Stresses” header (Figure 3). These represent the effective stiffnesses of cracked section with reinforcement in Y and Z direction respectively. With respect to gross-section the effective values will be as follows:
- Ey-eff/Ey-gross = 13872/36887 = 0.38
- Ez-eff/Ez-gross = 13164/36887 = 0.36

SOFiSTiK Command
+PROG AQB
HEAD Cracked Stiffness
PAGE UNII 5
NORM DC IS NDC IRC112 COUN IN
ECHO FULL NO
ECHO FORC YES
ECHO DESI YES
ECHO NSTR FULL
ECHO USEP NO
S NCS 1 NO 1 N -5000 MZ 10000 MY 5000
$Internal forces resulting at fixed end due to LC 1
NSTR KMOD ULTI
ENDComparison with non-linear analysis
Finally, the non-linear analysis of the model is performed to compare the results with the section check. The geometric non-linearity arising due to P-Delta effect is analysed using TH2 command in ASE module. For material nonlinearity the NSTR command is used. The analysis results are stored in LC 301. Using these results, the relative stiffnesses in Y and Z can be read from the interactive graphics. The values shown in Figure 4 at the fixed end are:
- Relative Stiffness EIy = 0.400
- Relative Stiffness EIz = 0.361
These values are very close to the one obtained using section check through AQB.
SOFiSTiK Command
+PROG ASE
HEAD TH2 With NSTR
SYST PROB TH2
NSTR KMOD S1 KSV SLD KSB SLD FMAX 0.85
LC 301
LCC 1
end
Key Takeaway / Learning Outcomes
This article presented two procedures in SOFiSTiK for obtaining the effective (cracked) stiffness of a reinforced concrete section: a section check with non-linear stress-strain curves in AQB, and a second-order non-linear analysis in ASE. For the 20 m M60 cantilever column studied, both give about 36-40% of the gross stiffness. The AQB check is quicker, since it needs only the governing forces from a linear analysis, while the ASE analysis also captures P-Delta effects and is preferable when second-order behaviour is significant.
Using gross stiffness overestimates member stiffness. In force-governed checks such as seismic analysis of indeterminate structures, this attracts higher moments and can lead to over-design, while deflections and drifts are underestimated.