User Tools

Site Tools


spin:esc203_fs2022

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision
Previous revision
spin:esc203_fs2022 [2022/03/14 11:22]
stadel [Lectures]
spin:esc203_fs2022 [2022/04/11 15:04] (current)
stadel [List of assignments]
Line 25: Line 25:
   - {{ :​spin:​degreve_reflection_refraction.pdf | Reflection and Refraction Tutorial}}   - {{ :​spin:​degreve_reflection_refraction.pdf | Reflection and Refraction Tutorial}}
  
 +21. Mar. 2022:
 +  - {{ :​spin:​teapot.zip | The famous UTAH Teapot triangultion (source: Stanford media lab)}}
 +  - {{ :​spin:​raytracingpart3.pdf | Ray Tracing - Part 3}}
 +
 +28. Mar. 2022:
 +  - {{ :​spin:​ray_tracing_-_part_4.pdf | Ray Tracing - Barycentric Triangle Coordinates and Vertex Normals}}
 +
 +4. Apr. 2022:
 +  - {{ :​spin:​multigridforellipticpdes_-_part_1.pdf | Multigrid for Elliptic PDEs - Part 1}}
 +
 +11. Apr. 2022:
 +  - {{ :​spin:​multigrid_method.pdf | Multigrid Method and "​subgrid"​ Boundary Conditions}}
  
 ====== Assignments ====== ====== Assignments ======
Line 73: Line 85:
 ====== List of assignments ====== ====== List of assignments ======
  
 +28. Mar. 2022: 
 +  - Fix the shadow bug by displacing the point +ve to the normal by a small epsilon.
 +  - Render transmission through the sphere as if it is made of glass. We have 3 sub-rays on the surface: Absorption (Phong), Reflection, and Transmission. ​
 +  - Render the Utah Teapot using the triangle intersection test. Choose your materials as you like (e.g., glass if you want). You could render a glass ball in front of an opaque teapot as well.
 +  - When you get the triangles rendered as factettes, then you are ready to compute the vertex normals using some sort of averaging of the face normals of the surrounding triangles. Then use the barycentric triangle coordinates to mix the 3 vertex normals to calculate an interpolated normal when calculating the absorption, reflection and transmission. This should give you a pretty smooth teapot. ​
  
 +11. Apr. 2022:
 +  - Get a multigrid (start with 2 grids) solver working for the simpler case of having the boundary conditions align with grid points at all levels of grid refinement. The BC should lie ontop of points in all of the grids. This requires a bit of care with the central plate that the power of 2 reductions of the grid remain on the endpoints of this plate. You can compare speed to SOR, but due to overheads, it my actually take longer with just 2 grids. With a deeper V-cycle this aspect should improve though. ​
spin/esc203_fs2022.1647253370.txt.gz ยท Last modified: 2022/03/14 11:22 by stadel