The purpose of this test case is to validate the plate/shell element solver in SkyCiv Structural 3D (S3D). The results were compared against a third-party analysis software. In this test case the following model parameters were used:
Analysis Type: Linear Static
Material: Structural Steel
Young’s Modulus: 200,000 MPa
Poisson’s Ratio: 0.27
Plate Type: Mindlin Plane Stress
Plate Size: 2m x 1m
Plate Thickness: 10 mm
Edge Boundary Supports: FFFRRR (3D Pin)
Pressure Load: 1 kPa normal to plate
Results – 8×4 Structured Mesh
Result
Location
SkyCiv
Third-Party 1
Difference 1
Third-Party 2
Difference 2
Max Displacement Z (mm)
Node 25 (center of plate)
-0.54879
-0.56405
-2.71%
-0.54879
0.00%
Max Von Mises Stress – Element (MPa)
Element 20 Top
4.63425
4.82904
-4.03%
4.63432
0.00%
Reaction Z (kN)
Node 21
0.12649
0.12961
-2.41%
0.12649
0.00%
Max Shear Force XZ – Element (kN/m)
Element 28
0.29475
0.33767
-12.71%
0.29475
0.00%
Results – 16×8 Structured Mesh
Result
Location
SkyCiv
Third-Party 1
Difference 1
Third-Party 2
Difference 2
Max Displacement Z (mm)
Node 120 (center of plate)
-0.55999
-0.56542
-0.96%
-0.55999
0.00%
Max Von Mises Stress – Element (MPa)
Element 72 Top
5.11488
5.17574
-1.18%
5.11487
0.00%
Reaction Z (kN)
Node 71
0.06218
0.06307
-1.41%
0.06218
0.00%
Max Shear Force XZ – Element (kN/m)
Element 120
0.35126
0.40086
-12.37%
0.35125
0.00%
Results – Unstructured Mesh
Result
Location
SkyCiv
Third-Party 1
Difference 1
Third-Party 2
Difference 2
Max Displacement Z (mm)
Node 182 (center of plate)
-0.55988
-0.56293
-0.54%
-0.55932
0.10%
Max Von Mises Stress – Element (MPa)
Element 170 Top
5.22850
5.25876
-0.58%
5.22487
0.07%
Reaction Z (kN)
Node 59
0.04218
0.04217
0.02%
0.04214
0.09%
Max Shear Force XZ – Element (kN/m)
Element 62
0.44973
0.49276
-8.73%
0.45013
-0.09%
Extra Considerations
This verification model was created and checked on 28 February 2020. Since this date, the plate solver and S3D software may have been further improved to achieved greater accuracy.
Plates are not exact elements like beam and frame elements and therefore the mesh plays a huge role in the results. Always try to use a structured mesh when it is possible to do so.
Results between software will never be exactly the same since different elements are used and the nature of plates are approximate