Cantilever 3

Train a CNN that predicts the stress contour in a cantilever beam with variable dimensions and a point load at the free end with variable magnitude and direction.

The input images represent information about the cantilever and loading using two channels:

• Black background with a white/gray line oriented at the specific angle and a brightness representing the load magnitude. Angle values follow the standard coordinate system and increase counterclockwise. An angle of 0 degrees points directly right, and an angle of 90 degrees points directly up.
• Black background with a white rectangle representing the length and height of the cantilever.

The output images are stress contours in the cantilever generated by FEA.

• Software: ANSYS (Mechanical APDL)
• Element type: PLANE182 (2D 4-node structural solid)
  • Thickness: 1
• Material: structural steel
  • Elastic modulus: 200 GPa
  • Poisson's ratio: 0.3

Cantilever 2

Train a CNN that predicts the stress contour in a fixed-size cantilever beam with a constant-magnitude point load at the end that varies in direction.

• Split script into three separate scripts.
• Renamed and added folders:
  • Renamed: Inputs → Train Inputs
  • Renamed: Outputs → Train Outputs
  • Added: Test Inputs
  • Added: Test Outputs

Cantilever 1

Train a CNN that predicts the stress contour in a fixed-size cantilever beam with a constant-magnitude point load at the end that varies in direction.