0% found this document useful (0 votes)
91 views11 pages

Frame Buckling

This document describes a buckling analysis of a one-story reinforced concrete frame structure in the elastic range using a linearized theory. It provides the geometry, loads, material properties, and steps to define member end stiffness coefficients and calculate the stiffness matrix, equivalent nodal loads, and solution of the system to determine nodal displacements and internal actions in the structure at the first analysis level, assuming an initial distribution of axial forces. The analysis uses an approximate method where the stiffness terms are corrected based on the assumed axial force distribution.

Uploaded by

Juan Diego
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
91 views11 pages

Frame Buckling

This document describes a buckling analysis of a one-story reinforced concrete frame structure in the elastic range using a linearized theory. It provides the geometry, loads, material properties, and steps to define member end stiffness coefficients and calculate the stiffness matrix, equivalent nodal loads, and solution of the system to determine nodal displacements and internal actions in the structure at the first analysis level, assuming an initial distribution of axial forces. The analysis uses an approximate method where the stiffness terms are corrected based on the assumed axial force distribution.

Uploaded by

Juan Diego
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 11

ONE-STORY R.C.

FRAME
BUCKLING ANALYSIS IN THE ELASTIC RANGE (Linearized theory)

See G.Toniolo and M. di Prisco, "CALCOLO STRUTTURALE I TELAI"

2 1
3 H p 3 H
B EJt C

EJp EJp

A D

FRAME GEOMETRICAL CHARACTERISTICS

Member AB n.1
BC n.2
CD n.3

b := 6000 mm

h := 4000 mm
1 3 4
Jp := ⋅ 300⋅ 400 mm
12

β := 2

Jt := β ⋅ Jp

LOADS

p := 300 kN/m=N/mm
H := 60000 N

I LEVEL ANALYSIS (APPROXIMATE METHOD)


We assume a plausible distribution of axial forces N and we carry out a second order analysis, with the stiffness terms
corrected by means of coefficients calculated on the basis of the assumed N distribution.

Example NAB= NCD=pb/2 = 900 kN


NBC= 0

DEFINITION OF THE MEMBER-END STIFFNESS COEFFICIENTS CORRECTIVE FUNCTIONS

Rotational stiffness corrective functions


κ
G1( κ ) := 1 +
30

κ
Gi( κ ) := 1 −
60

Translational stiffness corrective functions


κ
Gvi( κ ) := 1 + This is also the corrective coefficient that has to be applied to the shear generated by a
60 rotation of the beam end
κ
Gv ( κ ) := 1 +
10

Corrective function for the fixed-end moments (beam with a uniformly distributed load)
κ
G0( κ ) := 1 −
60

Moments of inertia vector

(
J := Jp Jt Jp )T
Member lengths vector

T
l := ( h b h )

Vector of the estimated axial loads N inherent with each member (tension has to be assumed positive)

T  −9 × 105 
p ⋅b p ⋅b   
N :=  − 0 − 
 2 2  N = 0 
 5
 −9 × 10 
i := 1 .. 3

MATERIAL MECHANICAL PROPERTIES

Rck := 55 MPa
Average elastic modulus (according to EC2)

1
3 4
Ecm := 9500⋅ ( 0.83⋅ Rck + 8) Ecm = 3.583 × 10 MPa

According to the C.N.R. 10025/98 code, the mechanical non-linearity can be considered by decreasing the elastic
modulus by means of a coefficient smaller than 1. If the coefficient is assumed equal to 1, a second order analysis with
an uncracked concrete section is carried out.
ψ := 0.46 ψ = 0.46
0.7 3
E := ψ ⋅ Ecm⋅ E = 9.614 × 10
1.2

Definition of ki:

κ :=
N⋅ l
i ( i)2  −0.936
κ = 0 
i E⋅ J  
i
 −0.936

STIFFNESS MATRIX
Restrained degrees of freedom:
1-rotation in B
2-rotation in C
3-horizontal displacement in C

The stiffness matrix K is a function of the vector k

 4⋅ E⋅ J1 4⋅ E⋅ J 2⋅ E⋅ J 6⋅ E⋅ J 
⋅ G1( κ ) + ⋅ G1( κ ) ⋅ Gi( κ ) ⋅ Gvi( κ )
2 2 1
 − 
( 1)
1 2 2 1
 l1 l
2
l
2 l
2 
 
 2⋅ E⋅ J 4⋅ E⋅ J 4⋅ E⋅ J 6⋅ E⋅ J 
⋅ Gi( κ ) ⋅ G1( κ ) + ⋅ G1( κ ) ⋅ Gvi( κ )
2 2 3 3
K ( κ ) :=  − 
(l3)
2 2 3 3
 l
2
l
2
l
3
2 
 
 6⋅ E⋅ J 6⋅ E⋅ J 12E⋅ J 12E⋅ J 
⋅ Gvi( κ ) ⋅ Gvi( κ ) ⋅ Gv ( κ ) + ⋅ Gv ( κ ) 
1 3 1 3
 − −
(l1)2 ( l 3) 2 (l1)3 (l3)3
1 3 1 3
 
 

 3.541× 1010 1.026× 1010 −5.679× 106 


 
K ( κ ) = 1.026 × 1010 3.541 × 1010 −5.679 × 106 

 
 6 6 3 
 −5.679 × 10 −5.679 × 10 5.229 × 10 
VECTOR OF THE EQUIVALENT NODAL LOADS

T  −9 × 108 
 ( 2) p ⋅ (l )
 p⋅ l 2 2   

⋅ G0( κ ) ⋅ G0( κ ) −H f = 9 × 108 

2
f := −
 12 2 12 2
  
 4
 −6 × 10 

SYSTEM SOLUTION (FIRST LEVEL)

 0.038 
u ( κ ) =  −0.034
−1
u ( κ ) := −K ( κ ) ⋅f Nodal displacements (First level)
 
 15.721

CALCULATION OF THE MOST SIGNIFICANT INTERNAL ACTIONS (FIRST LEVEL)

1) Axial loads

6⋅ E⋅ J 6⋅ E⋅ J p ⋅l
( 2) ( 2)
2 2 2 5
N := ⋅ Gvi κ ⋅ u ( κ ) + ⋅ Gvi κ ⋅ u ( κ ) − N = −8.8 × 10 Axial load - first column
( l 2) (l2)
1 2 1 2 2 2 1

6⋅ E⋅ J 12⋅ E⋅ J
( 1) ( )
2 1 1 5
N := − ⋅ H − ⋅ Gvi κ ⋅ u( κ ) + Gv κ ⋅ u ( κ ) N = −2.131 × 10 Axial load - beam
( l 1) ( l 1)
2 3 2 1 3 1 3 2

6⋅ E⋅ J 6⋅ E⋅ J p ⋅l
( 2) ( 2)
2 2 2 5
N := − ⋅ Gvi κ ⋅ u ( κ ) − ⋅ Gvi κ ⋅ u ( κ ) − N = −9.2 × 10 Axial load - second column
(l2) (l2)
3 2 1 2 2 2 3

6⋅ E⋅ J 12⋅ E⋅ J
( 3) ( )
1 3 3
N := ⋅H + ⋅ Gvi κ ⋅ u ( κ ) − Gv κ ⋅ u ( κ ) 5
N = −2.131 × 10 Axial load – beam (starting
(l3) ( l 3)
2 3 2 2 3 3 3
2
from column CD

2) Columns base moments

2⋅ E⋅ J 6⋅ E⋅ J
( 1) ( )
1 1 8
M := − ⋅ Gi κ ⋅ u ( κ ) + Gvi κ ⋅ u ( κ ) M = −2.054 × 10
(l1)
1 l 1 2 1 3 1
1

2⋅ E⋅ J 6⋅ E⋅ J
( 3) ( )
3 3 8
M := − ⋅ Gi κ ⋅ u ( κ ) + Gvi κ ⋅ u ( κ ) M = 3.535 × 10
(l3)
3 l 2 2 3 3 3
3
3) Column base shears

6⋅ E⋅ J 12⋅ E⋅ J
( 1) ( 1)
1 1 5
V := − ⋅ Gvi κ ⋅ u ( κ ) + ⋅ Gv κ ⋅ u ( κ ) V = −1.731 × 10
(l1) ( l 1)
1 2 1 3 3 1

6⋅ E⋅ J 12⋅ E⋅ J
( 3) ( 3)
3 3 5
V := − ⋅ Gvi κ ⋅ u ( κ ) + ⋅ Gv κ ⋅ u ( κ ) V = 2.331 × 10
(l3) ( l 3)
3 2 2 3 3 3

II LEVEL ANALYSIS (REFINED METHOD)


First of all we carry out a first order elastic analysis, finding a refined value of axial loads N. Then we carry out a second
order analysis, with the stiffness terms corrected by means of coefficients calculated on the basis of the refined N
distribution. The second level analysis does not depend upon the first level one.

We make use of the above introduced expressions, imposing ki=0 (all the corrective functions assume a value
equal to 1, hence obtaining a first order stiffness matrix)

 0
κ := 0 κ =  0
i  
 0
SYSTEM SOLUTION (FIRST ORDER ANALYSIS):

 0.037 
u ( κ ) =  −0.033
−1
u ( κ ) := −K ( κ ) ⋅f Nodal displacements (First order analysis)
 
 13.868

1) Axial loads (First order):

6⋅ E⋅ J 6⋅ E⋅ J p ⋅l
( 2) ( 2)
2 2 2 5
N := ⋅ Gvi κ ⋅ u ( κ ) + ⋅ Gvi κ ⋅ u ( κ ) − N = −8.822 × 10
( l 2) 2 (l2)2
1 1 2 2 1

6⋅ E⋅ J 12⋅ E⋅ J
( 1) ( )
2 1 1 5
N := − ⋅ H − ⋅ Gvi κ ⋅ u ( κ ) + Gv κ ⋅ u ( κ ) N = −2.125 × 10
( l 1) 2 ( l 1)
2 3 1 3 1 3 2

6⋅ E⋅ J 6⋅ E⋅ J p ⋅l
( 2) ( 2)
2 2 2 5
N := − ⋅ Gvi κ ⋅ u ( κ ) − ⋅ Gvi κ ⋅ u ( κ ) − N = −9.178 × 10
(l2)2 (l2)2
3 1 2 2 3

2) Columns base moments


2⋅ E⋅ J 6⋅ E⋅ J
( 1) ( )
1 1 8
M := − ⋅ Gi κ ⋅ u ( κ ) + Gvi κ ⋅ u ( κ ) M = −2.033 × 10
(l1)
1 l 1 2 1 3 1
1

2⋅ E⋅ J 6⋅ E⋅ J
( 3) ( )
3 3 8
M := − ⋅ Gi κ ⋅ u ( κ ) + G κ ⋅u(κ ) M = 3.367 × 10
2 vi 3
(l3)
3 l 2 3 3
3

3) Columns base shears

6⋅ E⋅ J 12⋅ E⋅ J
( 1) ( 1)
1 1 5
V := − ⋅ Gvi κ ⋅ u ( κ ) + ⋅ Gv κ ⋅ u ( κ ) V = −1.725 × 10
(l1) ( l 1)
1 2 1 3 3 1

6⋅ E⋅ J 12⋅ E⋅ J
( 3) ( 3)
3 3 5
V := − ⋅ Gvi κ ⋅ u ( κ ) + ⋅ Gv κ ⋅ u ( κ ) V = 2.325 × 10
(l3) ( l 3)
3 2 2 3 3 3

Redefinition of ki, based on the results of the first order analysis

κ :=
N⋅ l
i ( i)2  −0.918
κ =  −0.249
i E⋅ J  
i  −0.955

SYSTEM SOLUTION (SECOND LEVEL)

 0.038 
u ( κ ) =  −0.034
−1
u ( κ ) := −K ( κ ) ⋅f Nodal displacements (Second level)
 
 15.749
CALCULATION OF THE MOST SIGNIFICANT INTERNAL ACTIONS (SECOND LEVEL)

1) Axial loads

6⋅ E⋅ J 6⋅ E⋅ J p ⋅l
( 2) ( 2)
2 2 2 5
N := ⋅ Gvi κ ⋅ u( κ ) + ⋅ Gvi κ ⋅ u( κ ) − N = −8.8 × 10 Axial load - first column
( l 2) (l2)
1 2 1 2 2 2 1

6⋅ E⋅ J 12⋅ E⋅ J
( 1) ( )
2 1 1 5
N := − ⋅ H − ⋅ Gvi κ ⋅ u ( κ ) + Gv κ ⋅ u ( κ ) N = −2.148 × 10 Axial load - beam
( l 1) (l1)
2 3 2 1 3 1 3 2

6⋅ E⋅ J 6⋅ E⋅ J p ⋅l
( 2) ( 2)
2 2 2 5
N := − ⋅ Gvi κ ⋅ u( κ ) − ⋅ Gvi κ ⋅ u ( κ ) − N = −9.2 × 10 Axial load - second column
(l2) (l2)
3 2 1 2 2 2 3
6⋅ E⋅ J 12⋅ E⋅ J
( 3) ( )
1 3 3
N := ⋅H + ⋅ Gvi κ ⋅ u ( κ ) − Gv κ ⋅ u ( κ ) 5 Axial load – beam (starting
N = −2.148 × 10
( l 3) (l3)
2 3 2 2 3 3 3
2 from column CD

2) Columns base moments

2⋅ E⋅ J 6⋅ E⋅ J
( 1) ( )
1 1 8
M := − ⋅ Gi κ ⋅ u ( κ ) + G κ ⋅u(κ ) M = −2.076 × 10
2 vi 1
(l1)
1 l 1 3 1
1

2⋅ E⋅ J 6⋅ E⋅ J
( 3) ( )
3 3 8
M := − ⋅ Gi κ ⋅ u ( κ ) + Gvi κ ⋅ u ( κ ) M = 3.56 × 10
(l3)
3 l 2 2 3 3 3
3

3) Columns base shears

6⋅ E⋅ J 12⋅ E⋅ J
( 1) ( 1)
1 1 5
V := − ⋅ Gvi κ ⋅ u ( κ ) + ⋅ Gv κ ⋅ u ( κ ) V = −1.748 × 10
(l1) ( l 1)
1 2 1 3 3 1

6⋅ E⋅ J 12⋅ E⋅ J
( 3) ( 3)
3 3 5
V := − ⋅ Gvi κ ⋅ u ( κ ) + ⋅ Gv κ ⋅ u ( κ ) V = 2.348 × 10
(l3) ( l 3)
3 2 2 3 3 3

III LEVEL ANALYSIS (CORRECT METHOD)


We start from the II level solution and we keep updating the values of Ni, until the convergence criterion is satisfied.

κ :=
N⋅ l
i ( i)2 Redefinition of ki, based on the results of the II level analysis
 −0.915
κ =  −0.251
i E⋅ J  
i
 −0.957
Comparing the results, we find out that no more iterations are
 0.038  needed. If the convergence criterion is not satisfied, we evaluate the
u ( κ ) =  −0.034
−1
u ( κ ) := −K ( κ ) ⋅f new values of Ni, and we start again from the evaluation of ki.
 
 15.751
We have found the correct solution (III level)

CALCULATION OF THE MOST SIGNIFICANT INTERNAL ACTIONS (THIRD LEVEL)

1) Axial loads

6⋅ E⋅ J 6⋅ E⋅ J p ⋅l
( 2) ( 2)
2 2 2 5
N := ⋅ Gvi κ ⋅ u ( κ ) + ⋅ Gvi κ ⋅ u ( κ ) − N = −8.8 × 10 Axial load - first column
( l 2) (l2)
1 2 1 2 2 2 1
6⋅ E⋅ J 12⋅ E⋅ J
( 1) ( )
2 1 1 5
N := − ⋅ H − ⋅ Gvi κ ⋅ u ( κ ) + Gv κ ⋅ u ( κ ) N = −2.148 × 10 Axial load - beam
( 1) ( 1)
2 3 2 1 3 1 3 2
l l

6⋅ E⋅ J 6⋅ E⋅ J p ⋅l
( 2) ( 2)
2 2 2 5
N := − ⋅ Gvi κ ⋅ u ( κ ) − ⋅ Gvi κ ⋅ u ( κ ) − N = −9.2 × 10 Axial load - second column
( 2) ( 2)
3 2 1 2 2 2 3
l l

2) Columns base moments

2⋅ E⋅ J 6⋅ E⋅ J
( 1) ( )
1 1 8
M := − ⋅ Gi κ ⋅ u ( κ ) + G κ ⋅u(κ ) M = −2.076 × 10
2 vi 1
(l1)
1 l 1 3 1
1

2⋅ E⋅ J 6⋅ E⋅ J
( 3) ( )
3 3 8
M := − ⋅ Gi κ ⋅ u ( κ ) + G κ ⋅u(κ ) M = 3.561 × 10
2 vi 3
(l3)
3 l 2 3 3
3

3) Columns base shears

6⋅ E⋅ J 12⋅ E⋅ J
( 1) ( 1)
1 1 5
V := − ⋅ Gvi κ ⋅ u ( κ ) + ⋅ Gv κ ⋅ u ( κ ) V = −1.748 × 10
(l1) ( l 1)
1 2 1 3 3 1

6⋅ E⋅ J 12⋅ E⋅ J
( 3) ( 3)
3 3 5
V := − ⋅ Gvi κ ⋅ u ( κ ) + ⋅ Gv κ ⋅ u ( κ ) V = 2.348 × 10
(l3) (l3)
3 2 2 3 3 3

The calculated values coincide with those estimated in the II level.

CRITICAL BUCKLING LOAD CALCULATION (AT THE II LEVEL)

We keep H constant, while increasing p until the solution diverges.

Counter (a value greater than the critical j has to be provided as input):

jmax := 160
j := 1 .. jmax

Load multiplier:
j −1
γ := H := H γ =1
j 10 j 11

We make use of the above introduced expressions, imposing ki=0 (all the corrective functions assume a value equal to 1,
hence obtaining a first order stiffness matrix)
κ := 0 3 x jmax matrix
i, j
Data are built up as matrixes with 3 rows (3 problem unknowns) and j columns, where j stands for the step of the
evolving analysis. As a matter of fact, changing j, the distributed load (p) multiplier increases, until we get to the critical
multiplier (solution divergence).

γ ⋅p ⋅ l ( 2) 2 ⋅ G
( )
j
f := − 0 κ 2, j
1, j 12

 j ( 2) 
 γ ⋅ p ⋅ l 2 3 x jmax matrix
f
2, j
:=   ⋅ G0( κ 2 , j )
 12 

f := −H
3, j j

Nodal displacements vector for each j-th load combination:

〈j〉
u := −K κ
〈j〉 − 1 〈j〉
⋅f ( ) Nodal displacements (First order analysis)

N.B. to extract one column from a matrix, the command ctrl+6 can be used

First-order axial loads, for each j-th load:

6⋅ E⋅ J 6⋅ E⋅ J γ ⋅p ⋅l
( ) ( )
2 2 j 2 5
N := ⋅ Gvi κ ⋅u + ⋅ Gvi κ ⋅u − N = −8.822 × 10
1, j 2, j 1, j 2, j 2, j 1 , 11
( l 2) 2
( l 2) 2 2

6⋅ E⋅ J 12⋅ E⋅ J
( ) ( )
2 1 1 5
N := − ⋅ H − ⋅ Gvi κ ⋅u + Gv κ ⋅u N = −2.125 × 10
2, j 1, j 1, j 1, j 3, j 2 , 11
3 j
( l 1) 2
( l 1) 3

6⋅ E⋅ J 6⋅ E⋅ J γ ⋅p ⋅l
( ) ( )
2 2 j 2
N := − ⋅ Gvi κ ⋅u − ⋅ Gvi κ ⋅u − 5
3, j 2, j 1, j 2, j 2, j = −9.178 × 10
( 2) ( 2)
2 2 2 N
l l 3 , 11

Redefinition of ki, based on the results of the first-order analysis

κ :=
N
i, j
⋅ l ( i)2
i, j E⋅ J
i
At step 11 (load multiplier equal to 1) we get the same results of the previous analyses

0
0 -0.918
κ =
i , 11 1 -0.249
2 -0.955
SYSTEM SOLUTION (SECOND LEVEL)

〈j〉
u := −K κ ( )
〈j〉 − 1 〈j〉
⋅f Nodal displacements (Second level)

For j=11 we have:

 0.038 
( )
Same results of the previous analyses
〈11〉  〈11〉
= −0.034
24
u K κ = 4.331 × 10
 
 15.749

detK := K κ
j
( 〈j〉 ) detK
jmax
= −2.079 × 10
24

24
6×10

24
4×10

24
2×10
detK j
0

24
− 2×10

24
− 4×10
0 50 100 150
j
The critical load multiplier is:

jcritico := for j ∈ 2 .. jmax jcritico = 85


(
break if detK
j −1
≥ 0 ∧ detK ≤ 0
j )
jcritico ← j − 1

γcritico := γ γcritico = 8.4


jcritico

The following diagram shows the beam displacement versus the p load multiplier. The solution clearly diverges for a load
multiplier equal to the critical one (the determinant of the stiffness matrix tends to zero).

j := 1 .. jcritico
20

16

12
γj

0
3 3 3
0 2×10 4×10 6×10
u3 , j

You might also like