Adding tensor product wedges

[jlchan]

@Davknapp this is the draft I have for a tensor product wedge. The setup only involves changing the reference element constructor; everything else should remain the same. Here's an example of how I call it

using StartUpDG
N = 4
tensor_product_approximation_type = 
    TensorProductWedge(RefElemData(Tri(), N), RefElemData(Line(), N+4))
rd = RefElemData(Wedge(), tensor_product_approximation_type, N)

Trixi.jl specializes RefElemData for FDSBP types using DerivativeOperators from @ranocha's SummationByPartsOperators.jl package (see https://github.com/trixi-framework/Trixi.jl/blob/6e5512e8f4b8b2c3bf6811c6a3b0e548de4a91f6/src/solvers/dgmulti/sbp.jl#L66-L120). The intent is to use these with the TensorProductWedge type to construct discretization matrices.

There are two major changes needed for this PR:

• the face interpolation matrix needs to be set up - I think using Kronecker.jl (already used in StartUpDG.jl) would be the easiest way to construct these matrices. However, there are some performance issues associated with multiplication by these lazy Kronecker products due to allocations (Allocations using mul! MichielStock/Kronecker.jl#109).
• RefElemData currently assumes all differentiation matrices are the same type; however, it might make sense to allow the tensor product wedge to have differentiation matrices of different type (e.g., a sparse differentiation matrix in the vertical direction). This would be a breaking change v0.15 updates #58.

[codecov]

Codecov Report

Merging #59 (0adca35) into main (963eb3f) will decrease coverage by 2.10%.
The diff coverage is 0.00%.

@@            Coverage Diff             @@
##             main      #59      +/-   ##
==========================================
- Coverage   96.78%   94.69%   -2.10%     
==========================================
  Files          22       23       +1     
  Lines        2492     2547      +55     
==========================================
  Hits         2412     2412              
- Misses         80      135      +55     

Impacted Files	Coverage Δ
src/RefElemData_TensorProductWedge.jl	0.00% <0.00%> (ø)
src/StartUpDG.jl	100.00% <ø> (ø)

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[jlchan]

Closing in favor of #93