fixed systematic drift of grain structure along -[1,1,1]

correction of periodic Gauss kernel extension, inclusion of ndimage.gaussian_filter alternative (same runtime as FFT), proper output handling of grids <3D
This commit is contained in:
Philip Eisenlohr 2016-11-30 09:39:13 -05:00
parent 40de6910b8
commit cb95f3b244
1 changed files with 103 additions and 68 deletions

View File

@ -22,32 +22,46 @@ The final geometry is assembled by selecting at each voxel that grain index for
""", version = scriptID)
parser.add_option('-d', '--distance', dest='d', type='int', metavar='int',
parser.add_option('-d', '--distance',
dest = 'd',
type = 'float', metavar = 'float',
help = 'diffusion distance in voxels [%default]')
parser.add_option('-N', '--smooth', dest='N', type='int', metavar='int',
help='N for curvature flow [%default]')
parser.add_option('-r', '--renumber', dest='renumber', action='store_true',
help='renumber microstructure indices from 1...N [%default]')
parser.add_option('-i', '--immutable', action='extend', dest='immutable', metavar = '<int LIST>',
help='list of immutable microstructures')
parser.add_option('-N', '--iterations',
dest = 'N',
type = 'int', metavar = 'int',
help = 'curvature flow iterations [%default]')
parser.add_option('-i', '--immutable',
action = 'extend', dest = 'immutable', metavar = '<int LIST>',
help = 'list of immutable microstructure indices')
parser.add_option('-r', '--renumber',
dest = 'renumber', action='store_true',
help = 'output consecutive microstructure indices')
parser.add_option('--ndimage',
dest = 'ndimage', action='store_true',
help = 'use ndimage.gaussian_filter in lieu of explicit FFT')
parser.set_defaults(d = 1)
parser.set_defaults(N = 1)
parser.set_defaults(renumber = False)
parser.set_defaults(immutable = [])
parser.set_defaults(d = 1,
N = 1,
immutable = [],
renumber = False,
ndimage = False,
)
(options, filenames) = parser.parse_args()
options.immutable = map(int,options.immutable)
# --- loop over input files -------------------------------------------------------------------------
getInterfaceEnergy = lambda A,B: (A*B != 0)*(A != B)*1.0 # 1.0 if A & B are distinct & nonzero, 0.0 otherwise
struc = ndimage.generate_binary_structure(3,1) # 3D von Neumann neighborhood
# --- loop over input files -----------------------------------------------------------------------
if filenames == []: filenames = [None]
for name in filenames:
try:
table = damask.ASCIItable(name = name,
buffered = False, labeled = False)
try: table = damask.ASCIItable(name = name,
buffered = False,
labeled = False)
except: continue
damask.util.report(scriptName,name)
@ -56,11 +70,11 @@ for name in filenames:
table.head_read()
info,extra_header = table.head_getGeom()
damask.util.croak(['grid a b c: %s'%(' x '.join(map(str,info['grid']))),
'size x y z: %s'%(' x '.join(map(str,info['size']))),
'origin x y z: %s'%(' : '.join(map(str,info['origin']))),
'homogenization: %i'%info['homogenization'],
'microstructures: %i'%info['microstructures'],
damask.util.croak(['grid a b c: {}'.format(' x '.join(map(str,info['grid']))),
'size x y z: {}'.format(' x '.join(map(str,info['size']))),
'origin x y z: {}'.format(' : '.join(map(str,info['origin']))),
'homogenization: {}'.format(info['homogenization']),
'microstructures: {}'.format(info['microstructures']),
])
errors = []
@ -71,34 +85,30 @@ for name in filenames:
table.close(dismiss = True)
continue
# --- read data ------------------------------------------------------------------------------------
microstructure = np.tile(np.array(table.microstructure_read(info['grid']),'i').reshape(info['grid'],order='F'),
# --- read data -----------------------------------------------------------------------------------
microstructure = np.tile(table.microstructure_read(info['grid']).reshape(info['grid'],order='F'),
np.where(info['grid'] == 1, 2,1)) # make one copy along dimensions with grid == 1
grid = np.array(microstructure.shape)
#--- initialize support data -----------------------------------------------------------------------
# --- initialize support data ---------------------------------------------------------------------
# store a copy the initial microstructure to find locations of immutable indices
microstructure_original = np.copy(microstructure)
if not options.ndimage:
X,Y,Z = np.mgrid[0:grid[0],0:grid[1],0:grid[2]]
# Calculates gaussian weights for simulating 3d diffusion
gauss = np.exp(-(X*X + Y*Y + Z*Z)/(2.0*options.d*options.d))/math.pow(2.0*np.pi*options.d*options.d,1.5)
gauss[:,:,grid[2]/2::] = gauss[:,:,int(round(grid[2]/2.))-1::-1] # trying to cope with uneven (odd) grid size
gauss[:,grid[1]/2::,:] = gauss[:,int(round(grid[1]/2.))-1::-1,:]
gauss[grid[0]/2::,:,:] = gauss[int(round(grid[0]/2.))-1::-1,:,:]
gauss = np.fft.rfftn(gauss)
gauss = np.exp(-(X*X + Y*Y + Z*Z)/(2.0*options.d*options.d)) \
/math.pow(2.0*np.pi*options.d*options.d,(3.0 - np.count_nonzero(info['grid'] == 1))/2.)
getInterfaceEnergy = lambda A,B: (A*B != 0)*(A != B)*1.0 # 1.0 if A & B are distinct & nonzero, 0.0 otherwise
struc = ndimage.generate_binary_structure(3,1) # 3D von Neumann neighborhood
gauss[:,:,:grid[2]/2:-1] = gauss[:,:,1:(grid[2]+1)/2] # trying to cope with uneven (odd) grid size
gauss[:,:grid[1]/2:-1,:] = gauss[:,1:(grid[1]+1)/2,:]
gauss[:grid[0]/2:-1,:,:] = gauss[1:(grid[0]+1)/2,:,:]
gauss = np.fft.rfftn(gauss)
for smoothIter in range(options.N):
periodic_microstructure = np.tile(microstructure,(3,3,3))[grid[0]/2:-grid[0]/2,
grid[1]/2:-grid[1]/2,
grid[2]/2:-grid[2]/2] # periodically extend the microstructure
interfaceEnergy = np.zeros(microstructure.shape)
for i in (-1,0,1):
for j in (-1,0,1):
@ -112,30 +122,50 @@ for name in filenames:
periodic_interfaceEnergy = np.tile(interfaceEnergy,(3,3,3))[grid[0]/2:-grid[0]/2,
grid[1]/2:-grid[1]/2,
grid[2]/2:-grid[2]/2]
# transform bulk volume (i.e. where interfacial energy is zero)
# transform bulk volume (i.e. where interfacial energy remained zero), store index of closest boundary voxel
index = ndimage.morphology.distance_transform_edt(periodic_interfaceEnergy == 0.,
return_distances = False,
return_indices = True)
# want array index of nearest voxel on periodically extended boundary
periodic_bulkEnergy = periodic_interfaceEnergy[index[0],
index[1],
index[2]].reshape(2*grid) # fill bulk with energy of nearest interface
if options.ndimage:
periodic_diffusedEnergy = ndimage.gaussian_filter(
np.where(ndimage.morphology.binary_dilation(periodic_interfaceEnergy > 0.,
structure = struc,
iterations = int(round(options.d*2.)), # fat boundary
),
periodic_bulkEnergy, # ...and zero everywhere else
0.),
sigma = options.d)
else:
diffusedEnergy = np.fft.irfftn(np.fft.rfftn(
np.where(
ndimage.morphology.binary_dilation(interfaceEnergy > 0.,
structure = struc,
iterations = options.d/2 + 1), # fat boundary | PE: why 2d-1? I would argue for d/2 + 1
iterations = int(round(options.d*2.))),# fat boundary
periodic_bulkEnergy[grid[0]/2:-grid[0]/2, # retain filled energy on fat boundary...
grid[1]/2:-grid[1]/2,
grid[2]/2:-grid[2]/2], # ...and zero everywhere else
0.))*gauss)
periodic_diffusedEnergy = np.tile(diffusedEnergy,(3,3,3))[grid[0]/2:-grid[0]/2,
grid[1]/2:-grid[1]/2,
grid[2]/2:-grid[2]/2] # periodically extend the smoothed bulk energy
# transform voxels close to interface region | question PE: what motivates 1/2 (could be any small number, or)?
index = ndimage.morphology.distance_transform_edt(periodic_diffusedEnergy >= 0.5,
# transform voxels close to interface region
index = ndimage.morphology.distance_transform_edt(periodic_diffusedEnergy >= 0.5*np.amax(periodic_diffusedEnergy),
return_distances = False,
return_indices = True) # want index of closest bulk grain
periodic_microstructure = np.tile(microstructure,(3,3,3))[grid[0]/2:-grid[0]/2,
grid[1]/2:-grid[1]/2,
grid[2]/2:-grid[2]/2] # periodically extend the microstructure
microstructure = periodic_microstructure[index[0],
index[1],
index[2]].reshape(2*grid)[grid[0]/2:-grid[0]/2,
@ -146,10 +176,11 @@ for name in filenames:
# find locations where immutable microstructures have been or are now
for micro in options.immutable:
immutable += np.logical_or(microstructure == micro, microstructure_original == micro)
# undo any changes involving immutable microstructures
microstructure = np.where(immutable, microstructure_original,microstructure)
# --- renumber to sequence 1...Ngrains if requested ------------------------------------------------
# --- renumber to sequence 1...Ngrains if requested -----------------------------------------------
# http://stackoverflow.com/questions/10741346/np-frequency-counts-for-unique-values-in-an-array
if options.renumber:
@ -162,13 +193,14 @@ for name in filenames:
newInfo = {'microstructures': 0,}
newInfo['microstructures'] = microstructure.max()
# --- report ---------------------------------------------------------------------------------------
# --- report --------------------------------------------------------------------------------------
remarks = []
if (newInfo['microstructures'] != info['microstructures']): remarks.append('--> microstructures: %i'%newInfo['microstructures'])
if newInfo['microstructures'] != info['microstructures']:
remarks.append('--> microstructures: {}'.format(newInfo['microstructures']))
if remarks != []: damask.util.croak(remarks)
# --- write header ---------------------------------------------------------------------------------
# --- write header --------------------------------------------------------------------------------
table.labels_clear()
table.info_clear()
@ -185,8 +217,11 @@ for name in filenames:
# --- write microstructure information ------------------------------------------------------------
formatwidth = int(math.floor(math.log10(microstructure.max())+1))
table.data = microstructure.reshape((info['grid'][0],info['grid'][1]*info['grid'][2]),order='F').transpose()
table.data_writeArray('%%%ii'%(formatwidth),delimiter = ' ')
table.data = microstructure[::1 if info['grid'][0]>1 else 2,
::1 if info['grid'][1]>1 else 2,
::1 if info['grid'][2]>1 else 2,].\
reshape((info['grid'][0],info['grid'][1]*info['grid'][2]),order='F').transpose()
table.data_writeArray('%{}i'.format(formatwidth),delimiter = ' ')
# --- output finalization --------------------------------------------------------------------------