Minimization of Functional Majorant in a Posteriori Error Analysis Based on 𝐻(div) Multigrid-Preconditioned CG Method

Figure 2

Discrete solutions 𝑣 and y-component of the flux 𝑦 computed from Algorithm 2.2 on mesh triangulations 𝒯₂,𝒯₃,𝒯₄, and the exact solution 𝑢 and its flux y-component 𝜕𝑢/𝜕𝑦 (in the bottom row denoted as 𝒯_∞ triangulation).

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)