SAP2000 Verification
SAP2000 Verification
SAP2000 Verification
Integrated
Finite Element Analysis
and
Design of Structures
VERIFICATION MANUAL
COMPUTERS &
STRUCTURES
INC.
Version 6.1
Revised July 1997
COPYRIGHT
The computer program SAP2000 and all associated documentation are
proprietary and copyrighted products. Worldwide rights of ownership
rest with Computers and Structures, Inc. Unlicensed use of the program
or reproduction of the documentation in any form, without prior written
authorization from Computers and Structures, Inc., is explicitly prohibited.
Further information and copies of this documentation may be obtained
from:
DISCLAIMER
CONSIDERABLE TIME, EFFORT AND EXPENSE HAVE GONE
INTO THE DEVELOPMENT AND DOCUMENTATION OF
SAP2000. THE PROGRAM HAS BEEN THOROUGHLY TESTED
AND USED. IN USING THE PROGRAM, HOWEVER, THE USER
ACCEPTS AND UNDERSTANDS THAT NO WARRANTY IS EXPRESSED OR IMPLIED BY THE DEVELOPERS OR THE DISTRIBUTORS ON THE ACCURACY OR THE RELIABILITY OF
THE PROGRAM.
THE USER MUST EXPLICITLY UNDERSTAND THE ASSUMPTIONS OF THE PROGRAM AND MUST INDEPENDENTLY VERIFY THE RESULTS.
Table of Contents
Introduction
Example 1
Example 2
13
Example 3
17
Example 4
23
Example 5
29
Example 6
35
Example 7
39
Example 8
43
Example 9
47
Example 10
51
Example 11
59
Example 12
63
i
5
67
Example 14
71
Example 15
75
Example 16
79
Example 17
83
Example 18
87
Example 19
91
Example 20
ii
6
Introduction
This manual presents a set of example problems solved using the SAP2000 structural analysis program. These examples demonstrate many of the capabilities of the
SAP2000 program. For purposes of verification, key results from these examples
are compared with theoretical or published results from other computer programs,
where such are available. The verification problems cover each type of element and
include both static and dynamic examples.
For each example, this manual contains:
A short description of the problem
A list of significant SAP2000 options activated
A description of the input data used to create the model
A comparison of key results with theoretical results or results from other computer programs, if available
Some examples are solved using several different elements, mesh sizes and/or
boundary conditions. Key results from these different solutions are presented for
comparative purposes.
The data files and selected output files associated with the example problems are
provided in subdirectory EXAMPLES of the SAP2000 directory. For each example, the following files may be provided:
1
7
2
8
Example 1
Two-Dimensional Frame
Static and Dynamic Loads
Description
This is a seven-story two-dimensional frame structure, subjected to static lateral
and vertical loads and dynamic lateral loads due to earthquakes. The structure is
analyzed once using earthquake loads specified as a response spectrum, and once
using earthquake loads specified as a base acceleration time history. A solution to
this problem using another computer program is documented in Reference [1]. The
framing and the static loads are shown in Figure 1-1. The earthquake excitation is
shown in Figure 1-8. It is the N-S component of the 1940 El Centro earthquake. The
framing and the loads are all identical to the ones used in Reference [1].
3
9
Input Data
The computer model used is shown in Figures 1-1 and 1-2. Kip-inch units are used.
Two different input files are used. The first file is to analyze the structure for static
vertical and lateral loads and response spectrum dynamic loads. The input data file
for this analysis is FRAME. The second file is to repeat the dynamic analysis but
using base acceleration time-history as a loading. The input data file for this analysis is FRAMETH and the digitized base acceleration is given in file ELCENTRO.
This is shown in Figure 1-8.
The file FRAME is described first. Vertical loads input as Load Case 1 are specified
as span loading on beams. Static lateral loads input as Load Case 2 are specified as
joint loads. The lateral (Y) displacements of the columns at each story level are constrained together using a separate Diaphragm Constraint for each floor. Also,
masses are specified only in the lateral (Y) direction at each story level. These are
common modeling techniques used to reduce the size of the equation system and
are also utilized in the analysis reported in Reference [1]. The Diaphragm Constraints eliminate all axial deformations in the beam. This, and the absence of mass
specification in the vertical direction reduces the dynamic problem to seven natural
modes of vibration. All seven modes are included in the analysis.
It should also be noted that the AISC section properties in the database file SECTIONS.PRO are not used in this example and the required properties are explicitly
entered. This is intentional as most of the sections shown are older sections not in
the current AISC database.
The input file FRAMETH is identical to file FRAME for the structural model.
However, no static or response spectrum loads are specified. Instead the base acceleration is specified in the Y direction. The acceleration data is discretized in unequal time steps. The output sampling time used is 0.02 seconds and the response is
calculated for the first eight seconds. A damping value of 0.05 is used for all modes.
Comparison of Results
Reference [1] presents results only for the static lateral load analysis and the dynamic analysis. A comparison of key results for these analyses is presented in Figures 1-3, 1-4 and 1-5. The static results and the time periods are identical for the two
programs. The comparison is excellent for the response spectrum results and good
for the time-history results. Explicit time integration, not dependent on the size of
the time steps, is used in SAP2000.
For the response-spectrum results, the program of Reference [1] uses the square
root of the sum of the squares (SRSS) method of modal combination. SAP2000
4
10
Example 1
References
1. Static and Dynamic Analysis of Multistory Frame Structure Using DYNAMIC/EASE2, Engineering Analysis Corporation and Computers/Structures
International.
5
11
Figure 1-1
Two-dimensional Frame Example
6
12
Example 1
Figure 1-2
Two-dimensional Frame Example Model
7
13
Quantity
SAP2000
Reference [1]
Lateral Displacement
at Node 22
1.450764
1.450764
Axial Force
in Member 1
69.99
69.99
Moment in
Member 1
at Node 1
2324.68
2324.68
Figure 1-3
Comparison of Results for Static Lateral Loads
Mode
SAP2000
Reference [1]
1.2732
1.2732
0.4313
0.4313
0.2420
0.2420
0.1602
0.1602
0.1190
0.1190
0.0951
0.0951
0.0795
0.0795
Figure 1-4
Comparison of Results for Periods of Vibration
8
14
Example 1
SAP2000
Quantity
Reference [1]
Response
Spectrum
(CQC)
Response
Spectrum
(SRSS)
Time
History
Response
Spectrum
(SRSS)
Time
History
Lateral Displacement
at Node 22
5.431
5.437
5.486
5.438
5.46
Axial Force
in Member 1
261.5
261.7
263.0
261.8
258.0
Moment in
Member 1
at Node 1
9916
9864
9104
9868
8740
Figure 1-5
Comparison of Results for Dynamic Analyses
9
15
Figure 1-6
Deflection Due to Lateral Loads
Figure 1-7
Mode Shape 6
10
16
Example 1
Figure 1-8
El Centro Ground Acceleration Input
Figure 1-9
Lateral Displacement of Joint 22
11
17
18
Example 2
Bathe and Wilson Frame
Eigenvalue Problem
Description
This is a ten-bay, nine-story, two-dimensional frame structure solved in Reference
[1]. The framing and the material and section properties are shown in Figure 2-1.
The mass per unit length and other properties used are consistent with References
[1] and [2], to which the results are compared. The first three eigenvalues are computed.
Input Data
The computer model used is shown in Figure 2-1. Kip-foot units are used. Mass per
unit length of the members is specified. The program automatically computes the
joint masses to be used in the eigenvalue analysis.
The input data file for this example is FRAMEBW.
13
19
Comparison of Results
A comparison of the first three eigenvalues computed by SAP2000 with results
from References [1] and [2] is presented in Figure 2-2. The comparison is excellent.
References
1. Bathe, K. J. and Wilson, E. L.
Large Eigenvalue Problems in Dynamic Analysis, Journal of the Eng. Mech.
Div., ASCE, Vol. 98, No. EM6, Proc. Paper 9433, Dec. 1972.
2. Peterson, F. E.
EASE2, Elastic Analysis for Structural Engineering, Example Problem
Manual, Engineering Analysis Corporation, Berkeley, California, 1981.
14
20
Example 2
Figure 2-1
Bathe and Wilson Frame Example
15
21
Mode
SAP2000
Reference [1]
Reference [2]
0.589541
0.589541
0.589541
5.52696
5.52695
5.52696
Figure 2-2
Comparison of Eigenvalues
16
22
Example 3
Three-dimensional Frame
Dynamic Loads
Description
This is a two-story, three-dimensional frame structure with rigid diaphragm floors.
The problem is the same as the one solved in Reference [1]. The framing is shown
in Figures 3-1 and 3-2. The structure is doubly symmetric in plan, except that the
center of gravity at each story level is eccentric and is given by coordinates X = 38
feet and Y = 27 feet, represented in Figure 3-1 by joints 28 and 29.
Input Data
The computer model used is shown in Figures 3-1 and 3-2. Kip-foot units are used.
An additional joint is added to each story at the center of gravity, and all story mass
is given at these two joints.
17
23
Comparison of Results
A comparison of the SAP2000 results with Reference [1] results for the four natural
periods of vibration and the X-deflection at Joint 29 is presented in Figure 3-3. The
comparison is excellent.
Reference
1. Peterson, F. E.
EASE2, Elastic Analysis for Structural Engineering, Example Problem Manual, Engineering Analysis Corporation, Berkeley, California, 1981.
18
24
Example 3
Figure 3-1
Three-dimensional Frame Example: Dimensions
19
25
Figure 3-2
Three-dimensional Frame Example: Element Properties
20
26
Example 3
Quantity
SAP2000
Reference [1]
Period, Mode 1
0.2271
0.2271
Period, Mode 2
0.2156
0.2156
Period, Mode3
0.0733
0.0733
Period, Mode 4
0.0720
0.0720
X Deflection, Joint 29
0.0201
0.0201
Figure 3-3
Comparison of Results
21
27
28
Example 4
ASME Frame
Eigenvalue Problem
Description
This is a single-story, single-bay in each direction, three-dimensional frame structure made of 2-inch steel pipe segments and 2.75-inch steel cubes as shown in Figure 4-1. The frame is the same as modeled in References [1] and [2] and is Problem
1 from the ASME 1972 Program Verification and Qualification Library (Reference
[3]).
Input Data
The computer model used is shown in Figure 4-2. The pipe segments are modeled
using Frame elements. For consistency with Reference [1], masses are specified at
the nodes instead of using the Frame member mass per unit length specification and
additional nodal masses only for the solid cubes. Masses identical to those used in
23
29
Comparison of Output
A one-to-one comparison of SAP2000 results with References [1] and [2] results
for this problem is not possible. This is because both References [1] and [2] use the
Guyan reduction method to reduce the 42 dynamic degrees of freedom problem to
24. This introduces approximations into the solution. A comparison of the results is
presented in Figure 4-3 for the first 12 natural frequencies. Two sets of results are
presented for SAP2000: one using Ritz vector analysis, and one using eigenvector
analysis.
The comparison between SAP2000 and References [1] and [2] is good considering
the modeling differences between the different solutions. The SAP2000 eigenvector and Ritz-vector results are essentially the same for the first 11 modes, but begin
to differ in the higher modes. In general, only the eigenvectors represent the natural
modes of the structure. The Ritz vectors are a better basis for response-spectrum
and time-history analyses, but may not have the same frequencies and mode shapes
as the eigenvectors.
References
1. Peterson, F. E.
EASE2, Elastic Analysis for Structural Engineering, Example Problem Manual, Engineering Analysis Corporation, Berkeley, California, 1981.
2. DeSalvo, G. J. and Swanson, J. A.
ANSYS, Engineering Analysis System, Examples Manual, Swanson Analysis
Systems, Inc., Elizabeth, Pennsylvania, 1977.
3. Program Verification and Qualification Library, ASME Pressure Vessel and
Piping Division, Committee on Computer Technology, 1972.
24
30
Example 4
Figure 4-1
ASME Frame Example
25
31
Figure 4-2
ASME Frame Model
26
32
Example 4
Mode
SAP2000
Ritz Analysis
SAP2000
Eigen Analysis
114
114
112
119
119
116
141
141
138
222
222
218
399
399
404
422
422
423
450
450
452
550
550
554
774
774
736
10
800
799
762
11
909
909
853
12
955
946
894
Figure 4-3
Comparison of Modal Frequencies
27
33
34
Example 5
Three-dimensional Braced Frame
Dynamic Loads
Description
This is a three-story, L-shaped braced frame structure. The floors act as rigid diaphragms. The framing consists of four identical frames and the frame members
carry only axial loads. The problem is identical to the one solved in Reference [1].
The framing is shown in Figure 5-1.
Input Data
The computer model used for Frame 1 is shown in Figure 5-2. Kip-inch units are
used. The models for Frames 2, 3 and 4 are identical except that node numbers are
incremented by 12 and member numbers by 21, for each successive frame. It must
be noted that the common column between Frames 2 and 3 is modeled twice, once
for each frame. This is done for consistency with the modeling in Reference [1].
29
35
Comparison of Results
Key results from the SAP2000 analysis are compared with the Reference [1] solutions in Figure 5-3. The comparison is excellent.
Reference
1. Peterson, F. E.
EASE2, Elastic Analysis for Structural Engineering, Example Problem Manual, Engineering Analysis Corporation, Berkeley, California, 1981.
30
36
Example 5
Figure 5-1
Three-dimensional Frame Example
31
37
Figure 5-2
Three-dimensional Braced Frame Example Model of Frame 1
32
38
Example 5
Quantity
SAP2000
Reference [1]
Period, Mode 1
0.326887
0.326887
Period, Mode 2
0.320640
0.320640
Axial Force,
Member 1
279.47
279.48
Axial Force,
Member 4
194.50
194.50
Axial Force,
Member 5
120.52
120.52
Figure 5-3
Comparison of Results for Three-dimensional
Braced Frame Example
33
39
40
Example 6
Beam Steady-State Harmonic Loads
Description
This is a fixed-end beam in two dimensions subjected to a uniformly distributed
load which varies harmonically with respect to time. The beam is shown in Figure
6-1. The problem is the same as given in Reference [1]. The beam is solved twice,
once using the undamped, steady-state analysis option and once using the periodic
loading, time-history analysis option.
Input Data
The computer model used is shown in Figure 6-1. Pound-inch units are used.
For the steady-state option the frequency of the forcing function is provided in cycles per second. The input data file for this option is BEAM.
35
41
Comparison of Results
Reference [1] computes the deflection at the center of the beam to be -0.0541 sin
300t. The SAP2000 result for the amplitude of this deflection for the steady-state
option is -0.054535. It should be noted that Reference [1] uses only the first five
mode shapes of the beam for its computation. The SAP2000 results for the timehistory option are shown in Figure 6-3. The maximum amplitude being reported is
-0.05440. The comparison of results for both methods of analysis in SAP2000 with
the theoretical results is excellent.
Reference
1. Paz, M.
Structural Dynamics, Theory and Computations, Van Nostrand Reinhold,
1985.
36
42
Example 6
Figure 6-1
Beam Example
37
43
Figure 6-2
Time Variation of Loading
Figure 6-3
Center Span Displacement
38
44
Example 7
Two-dimensional Truss
Static Loads
Description
This is a two-dimensional truss structure. All members can carry only axial loads.
The truss is shown in Figure 7-1.
Input Data
The computer model used is shown in Figure 7-1. Truss members are modeled using frame elements with zero moments of inertia. Load case 1 is live load, and Load
case 2 is dead load. Load combination 1 is dead load plus live load.
The input data file for this example is TRUSS.
39
45
Comparison of Results
This example is included as a sample only. Other results for comparison are not
available.
40
46
Example 7
Figure 7-1
Two-dimensional Truss Example
41
47
48
Example 8
Three-dimensional Building
Dynamic Loads
Description
This is a two-story, three-dimensional framed building. The structure is shown in
Figure 8-1. The floors act as rigid diaphragms. The building is unsymmetrical and
is subjected to lateral dynamic loads along two horizontal axes at a 30o angle to the
building axes.
Input Data
The computer model used is shown in Figure 8-1. Kip-foot units are used. Joints 19
and 20 are specified at the center of gravity of Stories 1 and 2, respectively, and the
X- and Y-direction masses and the mass moment of inertia about the Z axis is defined at these joints only. All six modes of the structure are used for the dynamic
analysis. The response spectra are defined for two horizontal axes at a 30o angle to
the building axes.
43
49
Comparison of Results
This example is included as a sample only. Other results for comparison are not
available.
44
50
Example 8
Figure 8-1
Three-dimensional Building Example
45
51
52
Example 9
Patch Tests Prescribed
Displacements
Description
This is a rectangular plate with irregularly-shaped elements and subjected to prescribed displacements at the edges. The plate is shown in Figure 9-1. The location
of the inner nodes and the prescribed displacements are the same as suggested in
Reference [1]. The problem is solved using both Plane Stress elements and Shell
elements.
Input Data
The computer model used is shown in Figure 9-1. The prescribed displacements at
the edges are calculated as:
u = 10
-3
( x+
y
2
47
53
v = 10
-3
w = 10
( y+
( x
-3
x
2
+ xy + y
= 10
= 10
-3
-3
( y+
(-x-
x
2
y
2
Comparison of Results
The theoretical results for the problem are Sxx = Syy = 1333 and Sxy = 400 for the
membrane components; and Mxx = Myy = 1.111 x 10-7 and Mxy = 0.333 x 10-7 for the
plate bending components. These theoretical results are reproduced by SAP2000.
Reference
1. MacNeal, R. H. and Harder, R. C.
A Proposed Standard Set of Problems to Test Finite Element Accuracy, Finite
Elements in Analysis and Design 1 (1985), pp. - 20, North-Holland.
48
54
Example 9
Figure 9-1
Patch Test Example
49
55
56
Example 10
Straight Beam Static Loads
Description
This is a straight cantilever beam of proportions shown in Figure 10-1. The beam
has unit forces at the tip in the three orthogonal directions and a unit twist, each
modeled as a separate Load case. The beam is modeled using different mesh geometries as suggested in Reference [1]. The problem is solved using Shell elements,
nine-node Plane stress elements ,and Solid elements.
Input Data
Several mesh geometries are used and key results for these are presented. The following data files are provided:
STRBMSHL for the rectangular mesh of Shell elements.
STRBMPLN for the rectangular mesh of Plane elements.
51
57
Comparison of Results
The displacements at the tip in the direction of the load is compared for each type of
element and the different meshes with the theoretical results in Figures 10-4, 10-5,
and 10-6.
The results for the Shell element are good except for the in-plane shear results for
the irregular meshes and the results for the unit twist. The twist results are too stiff.
The element is too thick compared to the width for the thin plate twisting behavior
the Shell element is capable of modeling. The irregular mesh behavior in in-plane
shear can be improved by decreasing the skewness and aspect ratio of the elements
and using more elements.
52
58
Example 10
The results for the nine-node Plane element for all mesh geometries is excellent.
The four-node Plane element is too stiff in bending and a finer mesh would be required to accurately capture the bending behavior.
The results for the Solid element with a rectangular mesh and using incompatible
bending modes are good except for the case with the unit twist. To capture the twisting behavior accurately, more elements are needed across the beam section. The
Solid element model without incompatible bending modes is too stiff in the bending mode. A finer mesh along the length of the beam would give better results.
Reference
1. MacNeal, R. H. and Harder, R. C.
A Proposed Standard Set of Problems to Test Finite Element Accuracy, Finite
Elements in Analysis and Design 1 (1985), pp. 3-20, North-Holland.
53
59
Figure 10-1
Straight Beam Example - Shell Element Models
54
60
Example 10
Figure 10-2
Straight Beam Example - Plane Stress Element Model
Figure 10-3
Straight Beam Example - Solid Element
55
61
SAP2000
Load
Direction
Theoretical
Rectangular
Mesh
Trapezoidal
Mesh
Parallelogram
Mesh
Extension
3.000x10-5
3.000x10-5
3.000x10-5
3.000x10-5
Out-of-plane
Shear
0.4263
0.4266
0.4266
0.4321
In-plane Shear
0.1072
0.0221
0.0790
0.1081
Twist
0.00233
0.00233
0.00233
0.00321
Figure 10-4
Comparison of Tip Displacements Using Shell Elements
SAP2000
Load
Direction
Rectangular
Mesh
4-Node
Element
Rectangular
Mesh
9-Node
Element
Trapezoidal
Mesh
9-Node
Element
Parallelogram
Mesh
9-Node
Element
Extension
3.000x10-5
3.000x10-5
3.000x10-5
3.000x10-5
3.000x10-5
In-plane Shear
0.0101
0.1076
0.1063
0.1065
0.1081
Theoretical
Figure 10-5
Comparison of Tip Displacements Using Plane Stress Element
56
62
Example 10
SAP2000
Load
Direction
Rectangular
Mesh with
Incompatible
Bending Modes
Rectangular
Mesh without
Incompatible
Bending Modes
Extension
3.000x10-5
3.000x10-5
3.000x10-5
Out-of-plane Shear
0.4283
0.0109
0.4321
In-plane Shear
0.1072
0.0101
0.1081
Twist
0.00286
0.00286
0.00321
Theoretical
Figure 10-6
Comparison of Tip Displacements Using Solid Elements
57
63
64
Example 11
Curved Beam Static Loads
Description
This is a curved cantilever beam of proportions shown in Figure 11-1. The beam is
loaded with unit shears at the free end. The problem is the same as suggested in Reference [1]. The problem is solved using Shell elements and Plane elements.
Input Data
The input data file for the Shell element example is CRVBMSHL, and for the Plane
element example is CRVBMPLN.
Comparison of Results
The displacement at the free end in the direction of the load is compared with theoretical results (Reference [1]) for both element types in Figure 11-2.
For the Plane element, more elements along the length of the beam would give better results. The aspect ratio of the element is quite large in this model.
59
65
Reference
1. MacNeal, R. H. and Harder, R. C.
A Proposed Standard Set of Problems to Test Finite Element Accuracy, Finite
Elements in Analysis and Design 1 (1985), pp. 3-20, North-Holland.
60
66
Example 11
Figure 11-1
Curved Beam Example Shell Element Model
61
67
SAP2000
Load
Direction
In-plane
Shear
Theoretical
9-Node
Plane Stress
Element
Shell
Element
0.0775
0.0851
0.0873
0.4518
0.5022
Out-of-plane
Shear
Figure 11-2
Comparison of Tip Displacements
62
68
Example 12
Twisted Beam Static Loads
Description
This is a twisted cantilever beam of proportions shown in Figure 12-1. Unit loads
are applied at the free end. The problem is the same as suggested in Reference [1].
The problem is solved using Shell elements.
Input Data
The computer model used is shown in Figure 12-1. The input data file for this example is TWSBMSHL.
Comparison of Results
The displacements at the tip in the direction of the loads are compared with theoretical results (Reference [1]) in Figure 12-2. The comparison is excellent.
63
69
Reference
1. MacNeal, R. H. and Harder, R. C.
A Proposed Standard Set of Problems to Test Finite Element Accuracy, Finite
Elements in Analysis and Design 1 (1985), pp. 3-20, North-Holland.
64
70
Example 12
Figure 12-1
Twisted Beam Example
Load
Direction
SAP2000 Shell
Element
12x12 Mesh
Theoretical
In-plane Shear
0.005413
0.005424
Out-of-plane Shear
0.001770
0.001754
Figure 12-2
Comparison of Tip Deflections
65
71
72
Example 13
Beam On Elastic Foundation
Static Loads
Description
This is a simply supported beam on an elastic foundation. Half of the beam is modeled as shown in Figure 13-1. The geometry and the loads are the same as used in
Reference [1].
Input Data
The computer model used is shown in Figure 13-1. Pound-inch units are used.
Nine-node Plane elements with the plane stress option are used. Half of the beam is
modeled using 10 elements and symmetry is utilized to obtain the boundary conditions. Springs are used to model the elastic foundation.
The input data file for this example is BEAMONFN.
67
73
Comparison of Results
The transverse displacements along the center of the beam and the maximum bending stress (using the average of the values at the top and bottom fiber) are compared
with theoretical results (Reference [1]) in Figure 13-2. The comparison is good.
Reference
1. Peterson, F. E.
EASE2, Elastic Analysis for Structural Engineering, Example Problem Manual, Engineering Analysis Corporation, Berkeley, California, 1981.
68
74
Example 13
Figure 13-1
Beam on Elastic Foundation Example
69
75
Axial
Station
Transverse
Displacements
SAP2000
Theoretical
1.0458
1.0453
Maximum
Bending Stress
SAP2000
Theoretical
18029
18052
17751
17773
17183
17206
16304
16327
15082
15106
13476
13501
11436
11462
8902
8930
5810
5839
2089
2119
6
12
1.0336
1.0331
18
24
0.9973
0.9967
30
36
0.9373
0.9367
42
48
0.8546
0.8541
54
60
0.7507
0.7502
66
72
0.6275
0.6270
78
84
0.4874
0.4870
90
96
0.3335
0.3331
102
108
0.1695
0.1693
114
120
0.0002
0.0000
Figure 13-2
Comparison of Results
70
76
Example 14
Rectangular Plate Static Loads
Description
This is a rectangular plate as shown in Figure 14-1. This problem is solved using
two different aspect ratios: one square plate of 2 x 2, and another rectangular plate
of 2 x 10. Two Load cases are used: Load case 1 for a concentrated load at the center and Load case 2 for uniform load. Also, the problems are solved once with the
edges clamped, and once with the edges simply supported. The problem is modeled
using both Shell and Solid elements. The geometry, properties and loads used are
those suggested in Reference [1].
Input Data
The computer model using Shell elements is shown in Figure 14-1. For both types
of elements, a 6 x 6 mesh is used on a quarter of the plate, with symmetry conditions
applied to represent the remainder of the plate. Incompatible bending modes are included for the Solid element solutions.
71
77
Comparison of Results
The central deflection results for the various boundary conditions, loading conditions, and element types are compared in Figure 14-2 with the theoretical results.
The comparison is good. The thicker model using SOLID elements gives better results because of the improved aspect ratio.
Reference
1. MacNeal, R. H. and Harder, R. C.
A Proposed Standard Set of Problems to Test Finite Element Accuracy, Finite
Elements in Analysis and Design 1 (1985), pp. 3-20, North-Holland.
72
78
Example 14
Figure 14-1
Rectangular Plate Example
73
79
SAP2000
Loading
Boundary
Condition
Simply
Supported
Aspect
Ratio
(b/a)
0.0001
Thick
Shell
Element
1.0
4.061
3.175
4.084
4.062
5.0
12.92
9.37
12.91
12.97
1.0
1.29
0.83
1.30
1.26
5.0
2.60
1.83
2.61
2.56
1.0
11.77
8.76
11.89
11.60
5.0
17.74
6.40
15.97
16.96
1.0
5.76
3.44
5.87
5.60
5.0
7.80
1.78
6.39
7.23
Uniform
Clamped
Simply
Supported
Concentrated
Clamped
Figure 14-2
Comparison of Deflection at Center
74
80
Example 15
Cantilever Plate Eigenvalue Problem
Description
This is a square cantilever plate and the first five eigenvalues are computed. The
plate is shown in Figure 15-1.
Input Data
Three different mesh sizes are used: a 10 x 10 mesh, a 19 x 19 mesh, and a 40 x 40
mesh. The computer model used for the 19 x 19 mesh is shown in Figure 15-1.
Kip-inch units are used. Unit mass for the plate is specified to generate the mass
matrix.
The input data file is PLATE for the 19 x 19 mesh and LARGEPLT for the 40 x 40
mesh.
75
81
Comparison of Output
Reference [1] gives results for the first five natural frequencies of a square cantilever plate obtained by using the Ritz method with beam mode shapes. A comparison
of the SAP2000 results with the Reference [1] results is given in Figure 15-2.
The comparisons of the periods of vibration for the first, third and fourth modes are
excellent; however, the comparisons of the periods of vibration for the second and
fifth modes are not as good because the results of Reference [1] involve approximating plate mode shapes with beam mode shapes.
The SAP2000 results using three different mesh sizes are very close.
Reference
1. Harris, C. M. and Crede, C. E.
Shock and Vibration Handbook, McGraw-Hill, 1976.
76
82
Example 15
Figure 15-1
Cantilever Plate Example
77
83
SAP2000
Mode
40 x 40
19 x 19
10 x 10
MESH
MESH
MESH
0.01779
0.01781
0.01787
0.01790
0.00647
0.00648
0.00654
0.00732
0.00284
0.00285
0.00288
0.00292
0.00221
0.00223
0.00228
0.00228
0.00186
0.00187
0.00190
0.00201
Figure 15-2
Comparison of Natural Periods of Vibration
78
84
Reference [1]*
Example 16
Scordelis-Lo Roof Static Loads
Description
This is a reinforced concrete single span cylindrical roof as shown in Figure 16-1.
The roof is to be analyzed for gravity loads. The problem solved is the same as suggested in Reference [1].
Input Data
Two different mesh sizes are used for this problem: one, a 6 x 6 mesh used on a
quarter of the roof as shown in Figure 16-1; another, an 18 x 72 mesh used on half
the roof. Pound-foot units are used. Symmetry boundary conditions is used to represent the omitted parts of the structure. Unit weight is specified to apply the uniformly distributed loading.
The input data file for the 6 x 6 mesh is ROOF. The input data file for the 18 x 72
mesh is LARGEROF.
79
85
Comparison of Results
The theoretical vertical deflection at the center of the free edge is 0.3086, and Reference [1] suggests a value of 0.3024 for comparison of finite element behavior.
SAP2000 gives a value of 0.3068 for the 6 x 6 mesh and 0.3012 for the 18 x 72
mesh. The comparison is excellent.
A comparison of SAP2000 displacement and bending moment results using the 6 x
6 mesh, and theoretical results provided in Reference [2] using the theory presented
in Reference [3], is given in Figures 16-2 and 16-3. The theoretical results are measured from the figures given in Reference [2]. Nodal averages are used for the
SAP2000 bending moments shown. The comparison is good.
References
1. MacNeal, R. H. and Harder, R. L.
A Proposed Standard Set of Problems to Test Finite Element Accuracy, Finite
Elements in Analysis and Design 1 (1985), pp. 3-20, North-Holland.
2. Zienkiewicz, O. C.
The Finite Element Method, McGraw-Hill, 1977.
3. Scordelis, A. C. and Lo, K. S.
Computer Analysis of Cylindrical Shells, Journal of the American Concrete Institute, Vol. 61, May 1964.
80
86
Example 16
Figure 16-1
Scordelis-Lo Roof Example
81
87
Axial Deformations
at Support
Vertical Displacements
at Central Section
Location
SAP2000 6x6
Mesh
Theoretical
SAP2000 6x6
Mesh
Theoretical
0.0000
0.0004
0.046
0.045
6.67
0.0005
0.0009
0.031
0.027
13.33
0.0018
0.0020
-0.013
-0.018
20.00
0.0029
0.0030
-0.078
-0.082
26.67
0.0024
0.0021
-0.155
-0.155
33.33
-0.0017
-0.0016
-0.234
-0.241
40.00
-0.0118
-0.0120
-0.307
-0.309
Transverse Moments
at Central Section
Twisting Moments
at Support
Location
SAP2000 6x6
Mesh
Theoretical
SAP2000 6x6
Mesh
Theoretical
-2099
-2090
-91
6.67
-1978
-2000
-359
-380
13.33
-1614
-1620
-698
-670
20.00
-1045
-1000
-982
-1000
26.67
-408
-430
-1183
-1240
33.33
37
100
-1248
-1290
Figure 16-2
Comparison of Results
82
88
Example 17
Hemispherical Shell Static Loads
Description
This is a hemispherical shell loaded by point loads at the edge as shown in Figure
17-1. The problem solved is identical to the one suggested in Reference [1].
Input Data
The computer model used is shown in Figure 17-1. An 8 x 8 mesh of Shell elements
is used on a quarter of the hemispherical shell. Symmetry is used to define the
boundary conditions. A single restraint in the vertical direction is applied at the center of the free edge to provide stability.
The input data file for this example is SHELL.
Comparison of Results
The theoretical lower bound for the displacement under the load in the direction of
the load is 0.0924, where the hole at the center is not present. Reference [1] suggests
a value of 0.094 for comparison of results. The SAP2000 solution gives a value of
the displacement under the load of 0.0937. The comparison is excellent.
83
89
Reference
1. MacNeal, R. H. and Harder, R. C.
A Proposed Standard Set of Problems to Test Finite Element Accuracy, Finite
Elements in Analysis and Design 1 (1985), pp. 3-20, North-Holland.
84
90
Example 17
Figure 17-1
Hemispherical Shell Example
85
91
92
Example 18
Portal with P-delta
Description
These are two two-dimensional, one story, one bay, portal frames fixed at the base
as shown in Figure 18-1. One frame carries a central load of 1000 pounds and the
other carries an eccentric load of the same magnitude. The critical buckling load for
the frame in the symmetrical, non-sway mode is 6082 pounds (Pcr =
in the asymmetrical, sway mode is 1784 pounds (Pcr =
2 EI
.748
h2
2 EI
2.55
h2
) and
). The purpose of
this example is to verify the adequacy of the P-Delta algorithm in SAP2000 for
these type of problems. Theoretical results for these problems can be derived using
the stability function approach. Tables for stability functions can be obtained from
References [1].
87
93
Input Data
The two frames are modeled together in SAP2000 using Frame elements. The loading is applied as span loads to the beams of the structures. Pound-inch units are
used.
The theoretical results are computed assuming no axial deformation and no shear
deformation in the members. To duplicate this behavior a large axial area is specified for the elements and the shear areas are defaulted as zero (program interprets
this to mean that the shear deformation is to be ignored.) It is important to note that
the axial deformations of the members were not forced to be zero by restraining or
constraining the joints. Setting the axial deformation to be zero in this manner also
sets the axial load in the member to be zero. This then completely eliminates the PDelta effect from that member.
For the P-Delta analysis the default value of the relative displacement tolerance
(.001) is used and the maximum number of iterations is set at 10 even though the solution converges at a much lower number of iterations. All loads are in a single
Load case and this is used as the P-Delta load combination.
The input data for this example is PORTALPD.
Comparison of Results
The lateral displacement, joint rotation and member end-moment results for the two
load configurations are compared in Figure 18-2 with theoretical results and with
results for a model in which the P-Delta effect was ignored. The agreement between
the SAP2000 and theoretical results is excellent.
Theoretical values are obtained using the slope-deflection method of analysis by including the effect of axial loads in members. In this case the stiffness and carry-over
factor of each member, instead of being constant, are functions of the axial load in
the member. These functions are tabulated in Reference [1].
It is interesting to study the effect of P-Delta for these frames. In the symmetrical
frame (central load), the decrease in the stiffness of the columns due to compressive
axial loads causes a decrease in the beam end moments (top column moments) but
increases the bottom column moments. For the asymmetrical frame (eccentric
load), the same effect is seen in the column that carries the larger axial load. However, the column with the lesser axial load is now comparatively stiffer and attracts
a larger moment at each end when P-Delta effects are considered.
88
94
Example 18
Reference
1. Livesley, R. K., and Chandler, D. B.
Stability Functions for Structural Frameworks, Manchester University Press,
UK, 1956.
Figure 18-1
Portal Frame Example
89
95
Without
P-Delta
With P-Delta
QUANTITY
SAP2000
Theoretical
SAP2000 &
Theoretical
0.09178
0.09192
0.08620
4589.1
4606.6
4166.7
8260.4
8254.0
8333.3
Shear Force ( F )
128.5
128.6
125.0
Rotation Joint 2 ( 2
Without
P-Delta
With P-Delta
QUANTITY
SAP2000
Theoretical
SAP2000 &
Theoretical
Lateral Displacement
( )
1.894
1.893
1.385
Rotation Joint 6 ( 6
0.1014
0.1013
0.0924
Rotation Joint 8 ( 8
0.0367
0.0367
0.0369
2550.9
2544.9
2455.4
6183.6
6088.6
6919.6
4503.5
4456.9
3794.6
6124.9
6153.0
5580.4
Shear Force ( F )
101.6
101.4
93.75
Figure 18-2
Comparison of Results for a Planar Rigid Portal Frame
90
96
Example 19
Pounding of Two Planar Frames
Nonlinear Time-History Analysis
Description
A two-bay, seven-story plane frame is linked to a one bay four story plane frame using Nllink Gap elements. The structure experiences pounding due to ground motion. El Centro 1940 (N-S) record is used in the nonlinear time history analysis.
The geometry of the structure is shown in Figure 19-1.
Input Data
A Diaphragm constraint is applied to each floor of the two frames. Kip-inch units
are used.
The joints at column line 3 are connected to the corresponding joints at column line
4 by Nllink Gap elements. The local 1 axes of these elements are in the global X direction, and gap properties are specified for the local 1 direction. The opening dis91
97
Comparison of Results
This example is included as a sample only. A typical results plot is shown in Figure
19-2. It shows the variations of displacement of Column lines 3 and 4 and the link
force at Story level 4. It can be clearly seen that the link force is generated whenever
the two column lines move in phase and their separation is less than the specified
initial opening or if they move towards each other out of phase. For most part the
pounding has the effect of keeping the buildings in phase. For display purposes the
displacements are scaled up by a factor of 100.
92
98
Example 19
Figure 19-11
Geometry of Two Pounding Frames
93
99
Figure 19-2
Displacements of Pounding Columns and Pounding Force at Story 4
94
100
Example 20
Friction-Pendulum Base-Isolated 3D Frame
Nonlinear Time-History Analysis
Description
This is a two-story, three-dimensional frame with base isolation using friction-pendulum base isolators. The structure is subjected to earthquake motion in two perpendicular directions using the Loma Prieta acceleration records.
The geometry of the structure is shown in Figure 20-1.
Computer Model
The structure is modeled as a reinforced concrete frame with 9 column lines and 12
bays. Kip-inch units are used. The modulus of elasticity is taken to be 3000 ksi. The
self-weight of the concrete is taken as 150 pcf.
95
101
1E3
1E2
.04
.03
20
60
A single Load case is defined that models the self-weight of the structure. This will
be applied before the seismic analysis in order to generate the proper frictional response in the isolators. This self-weight must be applied as part of the time-history
analysis, as described below.
Seventeen Ritz vectors are requested for the modal analysis, using the two horizontal ground accelerations, the gravity load, and all the nonlinear deformation loads as
the starting load vectors. There are 27 nonlinear deformation loads, one for each of
the three translational degrees of freedom for the nine isolators. However, only
three of the 18 horizontal degrees of freedom are independent due to the diaphragm
constraint. The program will automatically discard 15 dependent starting load vectors. Thus the total number of independent starting load vectors is 15. By listing the
two ground accelerations as first, two Ritz vectors are generated for these two starting load vectors, and one for each of the remaining 13, for a total of 17 modes.
It is very important that the nonlinear deformation loads be used as starting load
vectors in order to capture nonlinear behavior. In this case, the axial load in each
isolator must be adequately modeled since the shear force depends strongly upon it.
Two time-history analyses are performed. The first history applies the self-weight
quasistatically. This history consists of 10 time steps of one second, for a total time
of 10 seconds. The load is increased linearly to full value for the first five seconds,
96
102
Example 20
and allowed to come to equilibrium for the remaining five seconds. A damping
value of 99% is used for all modes to prevent vibration.
The second history starts from the first history, and applies the seismic acceleration.
This history consists of 2000 time steps of 0.02 seconds, for a total time of 40 seconds. Zero damping is used for all modes since most energy dissipation is expected
to be due to the friction properties of the isolators, with little damage occuring in the
superstructure.
The input data file for this example is ISOLAT2 and the input time history files are
LP-TH0 and LP-TH90.
Comparison of Results
This example is included as a sample only. No results are available for comparison.
Typical plots are shown for the seismic time history. Figure 20-2 shows the time
variation of input, kinetic, potential (strain), damping, and frictional energies; these
energies do not start at zero since they continue from the self-weight time history.
Figure 20-3 shows the time variation of the displacements of the second floor at
Column line 1. Figure 20-4 shows the interaction diagram for the two shear forces
in the isolator beneath Column line 1. Figure 20-5 shows the force-displacement
plot for motion in the X direction of the isolator beneath Column line 1.
References
1. Zayas, V. and Low, S.
A Simple Pendulum Technique for Achieving Seismic Isolation, Earthquake
Spectra, Vol. 6, No. 2, Earthquake Engineering Research Institute, Oakland,
California,1990.
97
103
Figure 20-1
Model of Friction-Pendulum Base-Isolated 3-D Frame
98
104
Example 20
Figure 20-2
Variation of Energy during Seismic History
Figure 20-3
Horizontal Displacements at Top of Column 1 during Seismic History
99
105
Figure 20-4
Shear Force Interaction in Isolator under Column 1 during Seismic History
Figure 20-5
Shear Force vs. Deformation in Isolator under Column 1 during Seismic History
100
106
Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.
Alternative Proxies: