Plot3D Data

Plot3D Data

Plot3D output is a snapshot of five gas phase quantities over the whole mesh at a set of output times. FDS writes one .q file per mesh per time, plus one .xyz file per mesh holding the grid.

Enabling it in FDS

&DUMP DT_PL3D=30., WRITE_XYZ=.TRUE. /

WRITE_XYZ is required: the .q files hold no coordinates, so the .xyz files are the only source of the grid.

Reading a snapshot

import os
import pyfdstools as fds

workingDir = os.path.join(
    os.path.dirname(fds.__file__), 'examples', 'data', 'case001.zip')
chid = 'case001'

grid, values = fds.readPlot3Ddata(chid, workingDir, time=30.0)

print(grid.shape)     # (25, 11, 25, 3)
print(values.shape)   # (25, 11, 25, 5)

Note the argument order: chid comes before workingDir.

The output time nearest the one you ask for is used. For a multi-mesh case each mesh's own .q file for that time is read.

The five quantities

The last axis of values is fixed by the plot3D format:

Index	Quantity	Units
0	temperature	C
1	u-velocity	m/s
2	v-velocity	m/s
3	w-velocity	m/s
4	HRRPUV	kW/m3

temperature = values[:, :, :, 0]
hrrpuv = values[:, :, :, 4]

By default FDS writes those five; QUANTITIES on &DUMP can change what occupies each slot, in which case the table above no longer applies.

Taking a plane

x, z, frame = fds.findSliceLocation(grid, values[:, :, :, 0], axis=2, value=4.4)

fig, ax = fds.plotSlice(x, z, frame, axis=2,
                        clabel='Temperature (C)', qnty_mn=20, qnty_mx=200)
fig.savefig('pl3d_temperature.png')

plot3d=True returns all five fields at once instead:

x, z, T, U, V, W, HRR = fds.findSliceLocation(
    grid, values, axis=2, value=4.4, plot3d=True)

Which is what visualizePlot3D takes:

fds.visualizePlot3D(x, z, T, U, V, W, HRR, qnty_mn=20, qnty_mx=200)

plotSlice gives far more control; visualizePlot3D is a convenience wrapper that plots the temperature field only.

Printing the extents

fds.readPlot3Ddata(chid, workingDir, 30.0, verbose=True)

prints the coordinate and value ranges of each mesh as it is read, which is a quick sanity check that the case is what you think it is.

Reading one file

grid, gridHeader = fds.readXYZfile(xyzFile)
values, dataHeader = fds.readP3Dfile(qFile)

Both return flat arrays with the header separately; rearrangeGrid turns the grid into meshgrid form:

xGrid, yGrid, zGrid = fds.rearrangeGrid(grid)

Writing plot3D files

fds.writeP3Dfile(path, values)     # values.shape == (NX, NY, NZ, 5)
fds.writeXYZfile(path, grid)

Useful for feeding a derived field to smokeview, which will display any .q file named to match the case's .xyz files.

Errors

readPlot3Ddata raises FileNotFoundError if the case has no .xyz files, with a message reminding you to set WRITE_XYZ=.TRUE., or if a mesh has no .q file at any time.

The command line example

python extract_2d_slice_from_pl3d.py \
    --chid case001 --quantity TEMPERATURE \
    --axis 2 --value 4.4 --time 30 \
    --working_dir data/case001.zip