Hello, I am using a package JuAFEM for generating a three grids. After merging the

BridgeBot · February 19, 2021, 1:42am

Hello, I am using a package JuAFEM for generating a three grids. After merging the grids when I am trying to use the getfacesets function it does not show any faceset for the final merged grid, however it shows the facesets for individual grid. Below is the code:


using JuAFEM
using BlockArrays, SparseArrays, LinearAlgebra, Tensors

grid1 = generate_grid(QuadraticQuadrilateral,(10,10), Vec{2}((0.0,4.5)), Vec{2}((0.5,4.5)), Vec{2}((0.5,5.0)), Vec{2}((0.0,5.0)))
grid2 = generate_grid(QuadraticQuadrilateral,(10,10), Vec{2}((0.0,0.0)), Vec{2}((5.0,0.0)), Vec{2}((0.5,4.5)), Vec{2}((0.0,4.5)))
grid3 = generate_grid(QuadraticQuadrilateral,(10,10), Vec{2}((0.5,4.5)), Vec{2}((5.0,0.0)), Vec{2}((5.0,5.0)), Vec{2}((0.5,5.0)))

function merge_grids(grid1::Grid{dim,CellType}, grid2::Grid{dim,CellType}; tol=0.0001) where {N, dim, CellType &lt;: Cell{&lt;:Any, N}}
    cells′ = copy(grid1.cells)
    nodes′ = copy(grid1.nodes)
    nodemap = Dict{Int,Int}()
    next = getnnodes(grid1) + 1
    for (i2, n2) in enumerate(grid2.nodes)
        found = false
        for (i1, n1) in enumerate(grid1.nodes)
            if norm(n1.x - n2.x) &lt; tol
                nodemap[i2] = i1
                found = true
                break
            end
        end
        if !found
            push!(nodes′, n2)
            nodemap[i2] = next
            next += 1
        end
    end
    for c in grid2.cells
        t = ntuple(N) do i
            return nodemap[c.nodes[i]]
        end
        cell′ = CellType(t)
        push!(cells′, cell′)
    end
    return Grid(cells′, nodes′)
end

merge_grids(grid1::G, grid2::G, grids::G...) where G &lt;: Grid =
    merge_grids(merge_grids(grid1, grid2), grids...)

 grid = merge_grids(grid1,grid2, grid3)

getfacesets(grid)

Note that the original poster on Slack cannot see your response here on Discourse. Consider transcribing the appropriate answer back to Slack, or pinging the poster here on Discourse so they can follow this thread.
(Original message ) (More Info)