This page was generated from wta/adaptivity.ipynb.
Adaptivity¶
[1]:
from ngsolve import *
from ngsolve.webgui import Draw
Define the geometry by 2D Netgen-OpenCascade modeling (new in Netgen/NGSolve 2105):
[2]:
from netgen.occ import *
from netgen.webgui import Draw as DrawGeo
def MakeGeometryOCC():
base = Rectangle(1, 0.6).Face()
chip = MoveTo(0.5,0.15).Line(0.15,0.15).Line(-0.15,0.15).Line(-0.15,-0.15).Close().Face()
top = MoveTo(0.2,0.6).Rectangle(0.6,0.2).Face()
base -= chip
base.faces.name="base"
chip.faces.name="chip"
chip.faces.col=(1,0,0)
top.faces.name="top"
geo = Glue([base,chip,top])
geo.edges.name="default"
geo.edges.Min(Y).name="bot"
DrawGeo(geo)
return OCCGeometry(geo, dim=2)
geo = MakeGeometryOCC()
Define the geometry by curves (old-style segments geometry):
[3]:
# point numbers 0, 1, ... 11
# sub-domain numbers (1), (2), (3)
#
#
# 7-------------6
# | |
# | (2) |
# | |
# 3------4-------------5------2
# | |
# | 11 |
# | / \ |
# | 10 (3) 9 |
# | \ / (1) |
# | 8 |
# | |
# 0---------------------------1
#
def MakeGeometry():
from netgen.geom2d import SplineGeometry
geometry = SplineGeometry()
# point coordinates ...
pnts = [ (0,0), (1,0), (1,0.6), (0,0.6), \
(0.2,0.6), (0.8,0.6), (0.8,0.8), (0.2,0.8), \
(0.5,0.15), (0.65,0.3), (0.5,0.45), (0.35,0.3) ]
pnums = [geometry.AppendPoint(*p) for p in pnts]
# start-point, end-point, boundary-condition, left-domain, right-domain:
lines = [ (0,1,"bot",1,0), (1,2,"outer",1,0), (2,5,"outer",1,0), (5,4,"inner",1,2), (4,3,"outer",1,0), (3,0,"outer",1,0), \
(5,6,"outer",2,0), (6,7,"outer",2,0), (7,4,"outer",2,0), \
(8,9,"inner",3,1), (9,10,"inner",3,1), (10,11,"inner",3,1), (11,8,"inner",3,1) ]
for p1,p2,bc,left,right in lines:
geometry.Append(["line", pnums[p1], pnums[p2]], bc=bc, leftdomain=left, rightdomain=right)
geometry.SetMaterial(1,"base")
geometry.SetMaterial(2,"top")
geometry.SetMaterial(3,"chip")
return geometry
# geo = MakeGeometry()
# Draw(geo)
Piece-wise constant coefficients in sub-domains:
[4]:
mesh = Mesh(geo.GenerateMesh(maxh=0.2))
fes = H1(mesh, order=3, dirichlet="bot", autoupdate=True)
u, v = fes.TnT()
lam = CoefficientFunction([1, 1000, 10])
a = BilinearForm(fes)
a += lam*grad(u)*grad(v)*dx
# heat-source in inner subdomain
f = LinearForm(fes)
f += 1*v*dx(definedon="chip")
c = Preconditioner(a, type="multigrid", inverse="sparsecholesky")
gfu = GridFunction(fes, autoupdate=True)
Assemble and solve problem:
[5]:
def SolveBVP():
a.Assemble()
f.Assemble()
inv = CGSolver(a.mat, c.mat)
gfu.vec.data = inv * f.vec
SolveBVP()
Draw (gfu, mesh);
Gradient recovery error estimator: Interpolate finite element flux
\[q_h := I_h (\lambda \nabla u_h)\]
and take difference as element error indicator:
\[\eta_T := \tfrac{1}{\lambda} \| q_h - \lambda \nabla u_h \|_{L_2(T)}^2\]
[6]:
l = [] # l = list of estimated total error
space_flux = HDiv(mesh, order=2, autoupdate=True)
gf_flux = GridFunction(space_flux, "flux", autoupdate=True)
def CalcError():
# FEM-flux
flux = lam * grad(gfu)
# interpolate into H(div)
gf_flux.Set(flux)
# compute estimator:
err = 1/lam*(flux-gf_flux)*(flux-gf_flux)
eta2 = Integrate(err, mesh, VOL, element_wise=True)
l.append ((fes.ndof, sqrt(sum(eta2))))
print("ndof =", fes.ndof, " toterr =", sqrt(sum(eta2)))
# mark for refinement:
maxerr = max(eta2)
# marking with Python loop:
# for el in mesh.Elements():
# mesh.SetRefinementFlag(el, eta2[el.nr] > 0.25*maxerr)
# marking using numpy:
mesh.ngmesh.Elements2D().NumPy()["refine"] = \
eta2.NumPy() > 0.25*maxerr
CalcError()
ndof = 208 toterr = 0.006146664512048645
Adaptive loop:
[7]:
level = 0
while fes.ndof < 50000:
mesh.Refine()
SolveBVP()
CalcError()
level = level+1
if level%5 == 0:
Draw (gfu)
ndof = 355 toterr = 0.005412444050627307
ndof = 610 toterr = 0.0034764703269967627
ndof = 1057 toterr = 0.0024407198787836527
ndof = 1498 toterr = 0.0017725398350726652
ndof = 2176 toterr = 0.001175318481564381
ndof = 2977 toterr = 0.0007678238856744752
ndof = 3895 toterr = 0.0004992073011966752
ndof = 4711 toterr = 0.000328349661411465
ndof = 5509 toterr = 0.00022040792539907244
ndof = 6271 toterr = 0.00015347867373533968
ndof = 6934 toterr = 0.00011528702327142076
ndof = 7678 toterr = 9.010822245631904e-05
ndof = 8611 toterr = 6.964168425663449e-05
ndof = 9745 toterr = 5.3012198391045427e-05
ndof = 10642 toterr = 4.578385672462413e-05
ndof = 12292 toterr = 3.365421383991614e-05
ndof = 13735 toterr = 2.6279981723843566e-05
ndof = 15721 toterr = 2.0122030884461846e-05
ndof = 17944 toterr = 1.565278875918871e-05
ndof = 20470 toterr = 1.2331216391309937e-05
ndof = 23848 toterr = 9.22230353507417e-06
ndof = 27451 toterr = 7.2979670534441486e-06
ndof = 32353 toterr = 5.524333343051865e-06
ndof = 38077 toterr = 4.093132378390718e-06
ndof = 43858 toterr = 3.1725231314444495e-06
ndof = 51115 toterr = 2.4331738568182955e-06
[8]:
Draw (gfu);
[9]:
%matplotlib inline
import matplotlib.pyplot as plt
plt.yscale('log')
plt.xscale('log')
plt.xlabel("ndof")
plt.ylabel("H1 error-estimate")
ndof,err = zip(*l)
plt.plot(ndof,err, "-*")
plt.ion()
plt.show();