Software Verification

PROGRAM NAME: REVISION NO.:

SAP2000 0

EXAMPLE 2-007
SHELL HEMISPHERICAL SHELL STRUCTURE WITH STATIC LOADS PROBLEM DESCRIPTION In this example a hemispherical shell structure is analyzed for the effects of four 2 kip edge point loads alternating in sign at 90 intervals around the equator of the hemisphere. The deflections at the locations where the point loads are applied, in the direction of the point loads, are compared with published independent results. The geometry, properties and loading are as suggested in MacNeal and Harder 1985. The 0.04 foot thick hemispherical shell has a 10 foot radius. A hole is introduced at the top of the hemisphere, as shown in the figure on the following page, to avoid triangular elements at the top of the hemisphere. The top and bottom edges of the hemisphere are free. As suggested in MacNeal and Harder 1985, only a quarter of the hemispherical shell structure is modeled. Symmetrical boundary conditions are imposed on the side edges of the quarter hemisphere. An 8x8 mesh is used for the quarter hemispherical shell structure. To properly apply the symmetrical boundary conditions the local axes of each joint are rotated such that local axis 1 is in the same direction as a vector defined from the center point of the hemisphere to the point under consideration. In other words, local axis 1 is in the direction of a radial line to the point considered. Local axis 2 is horizontal and tangent to the hemispherical surface. With the joint local axes as described in the previous paragraph, the symmetry conditions are applied as follows. Joints 1, 10, 19, 28, 37, 46, 55, 64, and 73 are restrained in the U2 and R3 degrees of freedom. Joints 9, 18, 27, 36, 45, 54, 63, 72 and 81 are also restrained in the U2 and R3 degrees of freedom. In addition, a single vertical restraint is applied at the center of the bottom edge, at joint 5, to maintain stability of the structure.

EXAMPLE 2-007 - 1

Software Verification
PROGRAM NAME: REVISION NO.:

SAP2000 0

GEOMETRY, PROPERTIES AND LOADING

Material Properties E = 68,250,000 k/ft2 = 0.3 G =26,250,000 k/ft2 Section Properties Thickness = 0.04 ft 55 41 46 Sym 37 Z 25 28 17 19 7 9 10 5 1 1 F = 2.0 kips (1.0 kip on hemispherical quadrant) 2 2 3 3 4 4 5 Free 6 6 7 X 9 F = 2.0 kips 8 (1.0 kip on hemispherical quadrant) 8 Y
Radius o hemisp f here 10 '
18

Free 80 81 73 74 75 76 77 78 79 57 49 58 59 60 16 62 63 64 72 56

1 - Joint number 1 - Area object number 63 48 54 Sym 40 45 32 36 24 27 16 18

64

33

EXAMPLE 2-007 - 2

Software Verification
PROGRAM NAME: REVISION NO.:

SAP2000 0

TECHNICAL FEATURES OF SAP2000 TESTED Three-dimensional analysis using shell elements Joint local axes Joint force loads RESULTS COMPARISON MacNeal and Harder 1985 indicates that the theoretical lower bound for the displacement at the point load locations in the direction of the point load is 0.0924 feet for the condition where the hole at the center of the shell structure is not present. The reference further suggests a value of 0.0940 feet for comparison of results with the model where the center hole is present. The 0.0940 foot value is used in the comparison.

Shell Type Thin plate

Output Parameter Ux (jt 1) ft Uy (jt 9) ft

SAP2000 Independent 0.0939 -0.0939 0.0927 -0.0927 0.0940 -0.0940 0.0940 -0.0940

Percent Difference 0% 0% -1% -1%

Thick plate

Ux (jt 1) ft Uy (jt 9) ft

Note that for this example the thick plate solution gives less deformation than the thin plate solution. This occurs because the thick plate solution is stiffer in bending than the thin plate solution. Thus for shell models where the shear deformations are small compared to the bending deformations, the thick plate solution may have less deformation than the thin plate solution. COMPUTER FILES: Example 2-007-thick, Example 2-007-thin CONCLUSIONS The SAP2000 results for both the thin plate and thick plate options show an acceptable comparison with the independent results.

EXAMPLE 2-007 - 3