Direct Stiffness Method
Direct Stiffness Method
Direct Stiffness Method
a two-storey two-bay existing RC building. The first bay is infilled with one skin of masonry units
composed of clay bricks, which will be modeled as a pair of independent diagonal struts with an axial
stiffness of 5000KN/m. Cross section dimensions for all beams and columns are shown in figure 2. In
this case, assume a characteristic compressive strength of fck=30MPa and a Young modulus
Ec=4700(fck)0.5 for all of them. The seismic mass at each storey can be taken as 120t.
3,6
3,6
6 6
Figure 1. RC frame geometry and its elements and nodes number (dimensions in meters)
0,4
0,4
0,3
0,6
0,4
0,3
The building is located in a seismic area for which the corresponding design spectrum is defined as shown
in Figure 3. The fundamental period of structure can be computed using simplified formulation that
usually use in practical field as prescripted in ASCE Building Code for reinforced concrete frame is
By inputing the value of H equals to 7,2 m, then the fundamental period is,
Get the value of Sa(Ta) by plotting the in the design spectrum in Figure 3 below. For , the
Sa = 3 m/s2 .
3,5
3,0
2,5
2,0
Sa (m/s2)
1,5
1,0
0,5
0,0
0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,5
T (s)
where Sa(Ta) is the spectral acceleration at the fundamental period, M is the total seismic mass, and R is
the behavior factor (or ductility factor). For a RC infilled frame a value of R=3 can be assumed.
So, by inputing Sa = 3 m/s2 ; M = 2 x 120 t ; R = 3 and the Inelastic seismic base shear is,
The distribution of the inelastic base shear along the height of the building using the simplified profile
given with this equation,
∑
where Fi is the lateral force acting on the ith storey, zi is the height to the ith storey and Mi the
corresponding mass.
The direct stiffness method is used to perform linear static analysis of the frame structure subjected to
the seismic lateral load profile defined above. The structural nodes and element numbers are defined as in
Figure 4 below. The global and local reference system for each element are shown in Figure 5.
7 9 8 10 9
13 14
3,6
2 6 4
4 7 5 8 6
11 12
3,6
1 5 3
+y
1 2 3
+x
6 6
+y
i beam j
+x
j j
strut
column
+x
+y
+x
31°
i
90
+y i
°
+y
+y
+x
+x
U1 U4 U7
U3 U6 U9
U2 U5 U8
Figure 6. Determination of the degree of freedom (DoF) in global references before constrained the
bottom fixed-end.
The K matrix for each element should be obtained using the general K formulation for frame element
below. Note that the K element matrix is symmetric in diagonal.
From Figure 4 and 6, the mapping scheme is determined to arrange the global stiffness matrix with
constrained the nine degree of freedom of the bottom fixed end. The vector applied forces is arranged
from the lateral forces acting at the nodal 4 and 7 for their horizontal degree of freedom which are u10
and u19.
160 kN 7
80 kN 4
80 u10
0 u11
0 u12
0 u13
0 u14
0 u15
0 u16
0 u17
F= 0 u18
160 u19
0 u20
0 u21
0 u22
0 u23
0 u24
0 u25
0 u26
0 u27
Using the equations below, then we can obtain the displacements for each the degree of freedom.
{ } { } { }
{ } [ ]{ }
{ } [ ] { }
15,74
15,63 15,61
-0,23
-0,42 -0,18
0,215
-0,054 -0,12
8,95
8,89 8,88
-0,46
0,165 -0,92 -0,42
-0,042 -0,091
Nodal displacement
160kN
240kN
67,57 kNm
240kN
-61,90 kN -134,82 kN -36,28 kN
198,08 kNm 67,57 kNm
H (mm)
H (mm)
0 0 0
-50 -50 -50
-100 -100 -100
-150 -150 -150
-200 -200 -200
-0,5 0 -0,04 -0,02 0 0,02 0,04 -1,5 -1 -0,5 0 0,5 1 1,5
τ (N/mm2) σ (N/mm2) Pa (N/mm2)
Figure 9. The section force diagrams for shear, bending moment and axial force for exterior column due
to Vmax -68,91 kN, Mmax 67,57 kNm and N -125,76 kN
H (mm)
H (mm)
0 0 0
-50 -50 -50
-100 -100 -100
-150 -150 -150
-200 -200 -200
-250 -250 -250
-1 -0,5 0 -0,04 -0,02 0 0,02 0,04 -0,8-0,6-0,4-0,20,00,20,40,60,8
τ (N/mm2) σ (N/mm2) Pa (N/mm2)
Figure 10. The section force diagrams for shear, bending moment and axial force for interior column due
to Vmax -134,82 kN, Mmax 198,08 kNm and N 67,02 kN
Table 2. Total base shear and diplacement for 0.2Fb, 0.4Fb, 0.6Fb,0.8Fb, 1.0Fb
300
250
200
Base shear (kN)
150
100
50
0
0 5 10 15 20
Displacement (mm)
Figure 11. The total base shear versus top displacement curve.
In most of computer programs, inverting the stiffness matrix usually takes the greatest amount of
computational time. Different numerical methods exist to invert a matrix (e.g. Gauss, LU, LDLT
factorizations with direct/inverse substitution,etc.), most of them involving arithmetic operations on
different components of the matrix. Reducing the computational time and building efficient algorithms
can be done by exploiting the sparse nature of the stiffness matrix, and locating the off-diagonal non-zero
elements closer to the matrix diagonal. This is in order to reduce the “travel time”between components
when performing the arithmetic operations. This can be achieved with an adequate numbering of
structural nodes and elements before assembling K matrix.
A particular form of sparse scheme that is widely use in FEM codes is skyline storage. The skyline storage
format is a generalization of its widely used predecessor called the band storage scheme. A matrix stored
in accordance with the skyline format will be called a skymatrix for short. Only symmetric skymatrices
will bve considered here, and this match with the global stiffness matrix obtained in this problem case.
Since the stiffness matrix is symmetric, then we can define the envelope of K which the components are
non-zero. The envelope separates that entry from the rest of the upper triangle then the remaining zero
entries are conventionally removed. What is left constitute the skyline profile of skyline template of the
matrix. A sparse matrix that can be profitably stored in this form is called a skymatrix for brevity. Notice
that the skyline profile may include zero entries. During the factorization step discussed below these zero
entries will in general become nonzero, a phenomenon that receives the name fill-in.
7 9 8 10 9
13 14
3,6
2 6 4
4 7 5 8 6
11 12
3,6
1 5 3
+y
1 2 3
+x
6 6
Figure 12. The numbering of structural nodes and element of the RC frame.
u10 u11 u12 u13 u14 u15 u16 u17 u18 u19 u20 u21 u22 u23 u24 u25 u26 u27
1046010 -1029718 -4469,26 -8044,68 -3676,47 -2205,88 u10
1300092 30891,55 0 -10297,2 30891,55 0 -643574 0 -2205,88 -1323,53 u11
162180,6 0 -30891,6 61783,1 8044,675 0 9653,61 0 0 u12
2095040 0 0 -1029718 -3676,47 2205,882 0 -14125,1 0 -25425,1 u13
2311505 0 0 -10297,2 30891,55 2205,882 -1323,53 0 0 -1144132 0 u14
369173,1 0 -30891,6 61783,1 0 0 0 25425,15 0 30510,18 u15
1038657 0 0 0 0 0 0 0 0 -4469,26 -8044,68 u16
Kglobal = 1297445 -30891,6 0 0 0 0 0 0 0 -643574 0 u17
162180,6 0 0 0 0 0 0 8044,675 0 9653,61 u18
Symmetric 1037864 -2205,88 8044,675 -1029718 0 0 0 0 0 u19
655194,7 30891,55 0 -10297,2 30891,55 0 0 0 u20
142873,4 0 -30891,6 61783,1 0 0 0 u21
2077238 2205,882 25425,15 -1029718 0 0 u22
1166049 0 0 -10297,2 30891,55 u23
308152,8 0 -30891,6 61783,1 u24
1034188 0 8044,675 u25
653871,2 -30891,6 u26
142873,4 u27
The problem can be simplified using two different approach to consider the mechanism of horizontal
diplacement that assumed will happen in real building structure. The two approaches are the floor
diaphragm assumption and the shear frame assumption. A common floor diaphragm assumption in the
analysis of multistory buildings subjected to lateral loads is that the floor system undergoes no
deformation in its own plane. So instead of taking the all 18x18 degree of freedom, the global stiffness
matrix will be 14x14 by assuming that ust 1 = u10 = u13 = u16 and ust 2= u19 = u22 = u25. The global
stiffness, mass and external forces matrices for floor diaphragm assumption is shown after the Figure 14
below.
U19 U27
U21 U24
U20 U23 U26
U18
U10
U12 U15
U11 U14 U17
Floor diaphragm
U19
U10
shear frame
Figure 14. The floor diaphragm and the shear frame assumption.
ust 1 u11 u12 u14 u15 u17 u18 ust 2 u20 u21 u23 u24 u26 u27
60833,1 0 0 0 0 0 0 -30416,6 2205,882 -8044,68 -2205,88 -25425,1 0 -8044,68 ust 1 80 u10
1300092 30891,55 -10297,2 30891,55 0 0 -2205,88 -643574 0 -1323,53 0 0 0 u11 0 u11
162180,6 -30891,6 61783,1 0 0 8044,675 0 9653,61 0 0 0 0 u12 0 u12
2311505 0 -10297,2 30891,55 2205,882 -1323,53 0 -1144132 0 0 0 u14 0 u14
369173,1 -30891,6 61783,1 25425,15 0 0 0 30510,18 0 0 u15 0 u15
1297445 -30891,6 0 0 0 0 0 -643574 0 u17 0 u17
Kglobal = 162180,6 8044,675 0 0 0 0 0 9653,61 u18 F= 0 u18
30416,55 -2205,88 8044,675 2205,882 25425,15 0 8044,675 ust 2 160 u19
655194,7 30891,55 -10297,2 30891,55 0 0 u20 0 u20
142873,4 -30891,6 61783,1 0 0 u21 0 u21
1166049 0 -10297,2 30891,55 u23 0 u23
308152,8 -30891,6 61783,1 u24 0 u24
653871,2 -30891,6 u26 0 u26
142873,4 u27 0 u27
120 0 0 0 0 0 0 0 0 0 0 0 0 0 ust 1
0 0 0 0 0 0 0 0 0 0 0 0 0 u11
0 0 0 0 0 0 0 0 0 0 0 0 u12
0 0 0 0 0 0 0 0 0 0 0 u14
0 0 0 0 0 0 0 0 0 0 u15
0 0 0 0 0 0 0 0 0 u17
M= 0 0 0 0 0 0 0 0 u18
120 0 0 0 0 0 0 ust 2
0 0 0 0 0 0 u20
0 0 0 0 0 u21
0 0 0 0 u23
0 0 0 u24
0 0 u26
0 u27
The shear frame assumption in the analysis of multistory buildings subjected to lateral loads is that the
floor considered as rigid system undergoes no deformation in its own plane and the vertical and rotational
degree of freedom are neglected. So instead of taking the all 18x18 degree of freedom, the global stiffness
matrix will be only 2x2 by assuming that ust 1 = u10 = u13 = u16 and ust 2= u19 = u22 = u25. The
global stiffness, mass and external forces matrices for shear frame assumption is shown below.
ust 1 ust 2
Kglobal = 60833,1 -30416,6 ust 1 F= 80 ust 1
-30416,6 30416,55 ust 2 160 ust 2
120 0 ust 1
M= 0 120 ust 2
The comparisson of displacement obtained using three different approach are shown in the table below.
Dispalcement (mm)
n-story
Full DoF Floor diaphragm Shear Frame
8,95
8,91 7,89
1 8,89
8,88
15,74
2 15,63 15,66 13,15
15,61
Response-Spectrum analysis also can be used to solve the shear frame problem for a given structure with
mass m, stiffness k, and damping ξ, by utilized the design spectra for a spesific site. The procedure can be
used as follows:
1. Solve the eigen value and eigen vector of the coresponding K and M matrices:
Mode T Sa Sd (mm)
1 9,8396 0,64 2,9 9,98
2 25,7604 0,24 3,0 15,07
3,5
3,0
2,5
2,0
Sa (m/s2)
1,5
1,0
0,5
0,0
0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,5
T (s)
mi =120 t
U2 = 11,69 mm 137,35 kN 137,35 kN
137,35 kN
mi =120 t
90,25 kN 90,25 kN
U1=7,24 mm 220,12 kN
220,12 kN
220,12 kN
1313,9 kNm
Figure 16. Displacement, elastic forces, story shear profile and resistant forces
Table 5. Horizontal displacement, interstory-drift and elastic forces