Fig. (4)
Four FE models with growing computational effort due to mesh refinement: going from
(a)
to
(d)
models.