adds a new segmented pod rom helper class to handle joining per-zone pod models (a1e7656e) · Commits · dt-hydro / dt-hydro-cfd

dthydro_cfd/rom_cfd/pod_cfd_model.py

+132 −9

Original line number	Diff line number	Diff line
		@@ -16,6 +16,7 @@ import matplotlib.pyplot as plt
		import argparse
		import time
		import os
		from typing import List
		from scipy.interpolate import interp1d, RBFInterpolator
		import h5py
		try:
		@@ -28,9 +29,9 @@ v_i_name = "Velocity[i] (m/s)"
		v_j_name = "Velocity[j] (m/s)"
		v_k_name = "Velocity[k] (m/s)"
		p_name = "Static Pressure (Pa)"
		v_i_name = "Velocity in Rotating[i] (m/s)"
		v_j_name = "Velocity in Rotating[j] (m/s)"
		v_k_name = "Velocity in Rotating[k] (m/s)"
		#v_i_name = "Velocity in Rotating[i] (m/s)"
		#v_j_name = "Velocity in Rotating[j] (m/s)"
		#v_k_name = "Velocity in Rotating[k] (m/s)"


		def construct_pod_model(snapshot_h5file, data_name="Static Pressure (Pa)", n_pod_modes=5):
		@@ -91,6 +92,44 @@ class PodCfdModel(object):
		return self._xyz_points


		class SegmentedPodCfdModel(object):
		"""
		Contains several sub-POD ROM models that can be combined to form one output.
		Requires that each sub-POD ROM has the exact same inputs. Particularly useful when
		sharp boundaries in the flow or geometry exist in the full model: A different POD
		model can be fitted to the flow field in each region and later rejoined
		using this class.
		"""
		def __init__(self, snapshot_h5files: List[str], data_name: str="Static Pressure (Pa)", n_pod_modes: int=7):
		self._pod_cfd_models = []
		for snapshot_h5file in snapshot_h5files:
		self._pod_cfd_models.append(PodCfdModel(snapshot_h5file, data_name, n_pod_modes))

		def predict(self, vane_angle: float, min_vane_angle: float=4., max_vane_angle: float=14.):
		"""
		Evaluate the POD ROM.

		Args:
		vane_angle: turbine guide vane angle in degrees
		min_vane_angle: min vane angle (model only works
		between min and max CFD vane angle data)
		min_vane_angle: max vane angle (model only works
		between min and max CFD vane angle data)
		"""
		va = np.clip(vane_angle, min_vane_angle, max_vane_angle)
		seg_results = []
		for pod_model in self._pod_cfd_models:
		seg_results.append(pod_model.predict(va, min_vane_angle, max_vane_angle))
		return np.concatenate(seg_results, axis=0)

		@property
		def xyz_points(self):
		seg_xyz = []
		for pod_model in self._pod_cfd_models:
		seg_xyz.append(pod_model._xyz_points)
		return np.concatenate(seg_xyz, axis=0)


		def test_pod_cfd_model(use_plotly=True,
		snapshot_h5file="cfd_pod_snapshots.h5",
		n_pod_modes=7):
		@@ -155,6 +194,7 @@ def test_pod_cfd_model(use_plotly=True,
		print("Time to make plot: %0.4e (sec)" % (te-ts))

		else:
		# use matplotlib
		# create 3d plot figure of predicted pressure distribution
		ts = time.time()
		fig = plt.figure()
		@@ -169,6 +209,83 @@ def test_pod_cfd_model(use_plotly=True,
		print("Time to make plot: %0.4e (sec)" % (te-ts))


		def test_pod_segmented_cfd_model(snapshot_h5files, n_pod_modes=7):
		assert len(snapshot_h5files) > 1
		# create the reduced order model(s)
		pod_model = SegmentedPodCfdModel(
		snapshot_h5files, data_name=p_name,
		n_pod_modes=n_pod_modes)
		xyz_points = pod_model.xyz_points

		pod_model_vx = SegmentedPodCfdModel(
		snapshot_h5files, data_name=v_i_name, n_pod_modes=n_pod_modes)
		pod_model_vy = SegmentedPodCfdModel(
		snapshot_h5files, data_name=v_j_name, n_pod_modes=n_pod_modes)
		pod_model_vz = SegmentedPodCfdModel(
		snapshot_h5files, data_name=v_k_name, n_pod_modes=n_pod_modes)

		# evaluate the pod model at an vane angle not in training set
		ts = time.time()
		current_vane_angle = 12.5
		pred_p = pod_model.predict(current_vane_angle)
		pred_vx = pod_model_vx.predict(current_vane_angle)
		pred_vy = pod_model_vy.predict(current_vane_angle)
		pred_vz = pod_model_vz.predict(current_vane_angle)
		pred_vmag = np.sqrt(pred_vx2 + pred_vy2 + pred_vz**2)
		te = time.time()
		print("Time to eval POD ROM: %0.4e (sec)" % (te-ts))

		X = np.asarray(xyz_points[:, 0], dtype=np.float32)
		Y = np.asarray(xyz_points[:, 1], dtype=np.float32)
		Z = np.asarray(xyz_points[:, 2], dtype=np.float32)
		values = np.asarray(pred_p, dtype=np.float32)

		# make plots
		import plotly.graph_objects as go
		import plotly
		ts = time.time()
		fig_p = go.Figure(data=go.Scatter3d(
		x=X.flatten(),
		y=Y.flatten(),
		z=Z.flatten(),
		mode='markers',
		marker=dict(size=4, color=values, opacity=0.95,
		colorbar=dict(thickness=5, title="Static Pressure (Pa)")),
		customdata=values,
		hovertemplate=('%{customdata}%'),
		))
		fig_p.show()

		fig_v = go.Figure(data=go.Cone(
		x=X.flatten(),
		y=Y.flatten(),
		z=Z.flatten(),
		u=pred_vx,
		v=pred_vy,
		w=pred_vz,
		sizemode="absolute",
		sizeref=150,
		colorbar=dict(thickness=8, title="Velocity (m/s)"),
		))
		fig_v.show()
		# fig_vs = go.Figure(data=go.Streamtube(
		# x=X.flatten(),
		# y=Y.flatten(),
		# z=Z.flatten(),
		# u=pred_vx,
		# v=pred_vy,
		# w=pred_vz,
		# starts= dict(
		# x=[3.32,],
		# y=[1.0,],
		# z=[-1.75,]
		# ),
		# ))
		# fig_vs.show()
		te = time.time()
		print("Time to make plot: %0.4e (sec)" % (te-ts))


		def test_dynamic_pod_model(snapshot_h5file="cfd_pod_snapshots.h5", n_pod_modes=7):
		# create the reduced order model(s)
		pod_model = PodCfdModel(
		@@ -258,14 +375,20 @@ def test_dynamic_pod_model(snapshot_h5file="cfd_pod_snapshots.h5", n_pod_modes=7

		if __name__ == "__main__":
		parser = argparse.ArgumentParser()
		parser.add_argument("-i", help="Input POD snapshot file.", type=str, default="cfd_pod_snapshots.h5")
		parser.add_argument("-i", help="Input POD snapshot file(s).", type=str, nargs='+', default=["cfd_pod_snapshots.h5"])
		parser.add_argument("--dynamic", help="Run in dynamic test mode with moving vane angle", type=int, default=0)
		parser.add_argument("--n_pod_modes", help="Run in dynamic test mode with moving vane angle", type=int, default=7)
		args = parser.parse_args()
		if len(args.i) == 1:
		if bool(args.dynamic):
		test_dynamic_pod_model(
		args.i, n_pod_modes=args.n_pod_modes)
		snapshot_h5file=args.i[0],
		n_pod_modes=args.n_pod_modes)
		else:
		test_pod_cfd_model(
		use_plotly=True, snapshot_h5file=args.i,
		use_plotly=True, snapshot_h5file=args.i[0],
		n_pod_modes=args.n_pod_modes)
		else:
		test_pod_segmented_cfd_model(
		snapshot_h5files=args.i,
		n_pod_modes=args.n_pod_modes)

dthydro_cfd/rom_cfd/preprocess_cfd.py

+66 −37

Original line number	Diff line number	Diff line
		@@ -37,46 +37,65 @@ run_parameters = [
		]

		# NEW CFD DATA FILES
		run_dir = "./alder_u12/"
		# run_dir = "./alder_u12_fixed/"
		run_dir = "./alder_u12_fixed/"
		run_csv_files = [
		"alpha-01_runner.csv",
		"alpha-02_runner.csv",
		"alpha-03_runner.csv",
		"alpha-04_runner.csv",
		"alpha-05_runner.csv",
		"alpha-06_runner.csv",
		"alpha-07_runner.csv",
		"alpha-08_runner.csv",
		"alpha-09_runner.csv",
		# "alpha-01_runner.csv",
		# "alpha-02_runner.csv",
		# "alpha-03_runner.csv",
		# "alpha-04_runner.csv",
		# "alpha-05_runner.csv",
		# "alpha-06_runner.csv",
		# "alpha-07_runner.csv",
		# "alpha-08_runner.csv",
		# "alpha-09_runner.csv",
		# "alpha-10_runner.csv",
		"alpha-9_runner.csv",
		"alpha-10_runner.csv",
		]
		run_csv_files = [
		"alpha-01_spiral-casing-and-draft-tube.csv",
		"alpha-02_spiral-casing-and-draft-tube.csv",
		"alpha-03_spiral-casing-and-draft-tube.csv",
		"alpha-04_spiral-casing-and-draft-tube.csv",
		"alpha-05_spiral-casing-and-draft-tube.csv",
		"alpha-06_spiral-casing-and-draft-tube.csv",
		"alpha-07_spiral-casing-and-draft-tube.csv",
		"alpha-08_spiral-casing-and-draft-tube.csv",
		"alpha-09_spiral-casing-and-draft-tube.csv",
		"alpha-10_spiral-casing-and-draft-tube.csv",
		]
		# run_csv_files = [
		# # "alpha-01_spiral-casing-and-draft-tube.csv",
		# # "alpha-02_spiral-casing-and-draft-tube.csv",
		# # "alpha-03_spiral-casing-and-draft-tube.csv",
		# # "alpha-04_spiral-casing-and-draft-tube.csv",
		# # "alpha-05_spiral-casing-and-draft-tube.csv",
		# # "alpha-06_spiral-casing-and-draft-tube.csv",
		# # "alpha-07_spiral-casing-and-draft-tube.csv",
		# # "alpha-08_spiral-casing-and-draft-tube.csv",
		# # "alpha-09_spiral-casing-and-draft-tube.csv",
		# # "alpha-10_spiral-casing-and-draft-tube.csv",
		# "alpha-9_casing-draft_tube.csv",
		# "alpha-10_casing-draft_tube.csv",
		# ]
		# run_csv_files = [
		# # "alpha-01_distributor.csv",
		# # "alpha-02_distributor.csv",
		# # "alpha-03_distributor.csv",
		# # "alpha-04_distributor.csv",
		# # "alpha-05_distributor.csv",
		# # "alpha-06_distributor.csv",
		# # "alpha-07_distributor.csv",
		# # "alpha-08_distributor.csv",
		# # "alpha-09_distributor.csv",
		# # "alpha-10_distributor.csv",
		# "alpha-9_distributor.csv",
		# "alpha-10_distributor.csv",
		# ]
		run_parameters = [
		[1.3875],
		[2.775],
		[4.1625],
		[5.55],
		[6.9375],
		[8.325],
		[9.7125],
		[11.1],
		# [1.3875],
		# [2.775],
		# [4.1625],
		# [5.55],
		# [6.9375],
		# [8.325],
		# [9.7125],
		# [11.1],
		[12.4875],
		[13.875],
		]


		def plot_cfd_vel_field(mesh_x, mesh_y, mesh_z, v_i, v_j, v_k):
		def plot_cfd_vel_field(mesh_x, mesh_y, mesh_z, v_i, v_j, v_k, cbar_label=""):
		import plotly.graph_objects as go
		import plotly

		@@ -96,8 +115,8 @@ def plot_cfd_vel_field(mesh_x, mesh_y, mesh_z, v_i, v_j, v_k):
		v=v_j[subsample_idxs],
		w=v_k[subsample_idxs],
		sizemode="absolute",
		sizeref=150,
		colorbar=dict(thickness=8, title="CFD Velocity (m/s)"),
		sizeref=25,
		colorbar=dict(thickness=8, title=cbar_label+"CFD Velocity (m/s)"),
		))
		fig_v.show()
		fig_p = go.Figure(data=go.Scatter3d(
		@@ -106,7 +125,7 @@ def plot_cfd_vel_field(mesh_x, mesh_y, mesh_z, v_i, v_j, v_k):
		z=mesh_z[subsample_idxs].flatten(),
		mode='markers',
		marker=dict(size=4, color=v_mag[subsample_idxs], opacity=0.95,
		colorbar=dict(thickness=5, title="CFD Vel Magnitude (m/s)")),
		colorbar=dict(thickness=5, title=cbar_label+"CFD Vel Magnitude (m/s)")),
		customdata=v_mag[subsample_idxs],
		hovertemplate=('%{customdata}%'),
		))
		@@ -168,8 +187,12 @@ def conv2hdf(csv_fname, outfile_name="fixed_cfd.h5", case_params=[], dset_names=
		v_i = hdf["/Velocity[i] (m/s)"][:]
		v_j = hdf["/Velocity[j] (m/s)"][:]
		v_k = hdf["/Velocity[k] (m/s)"][:]
		vr_i = hdf["Velocity in Rotating[i] (m/s)"][:]
		vr_j = hdf["Velocity in Rotating[j] (m/s)"][:]
		vr_k = hdf["Velocity in Rotating[k] (m/s)"][:]
		p_s = hdf["/Static Pressure (Pa)"][:]
		plot_cfd_vel_field(xx, yy, zz, v_i, v_j, v_k)
		plot_cfd_vel_field(xx, yy, zz, vr_i, vr_j, vr_k, cbar_label="Rotating Ref ")
		plot_cfd_pres_field(xx, yy, zz, p_s)

		# clean up
		@@ -179,7 +202,8 @@ def conv2hdf(csv_fname, outfile_name="fixed_cfd.h5", case_params=[], dset_names=
		def reinterpolate_cfd(
		h5_fname_base, h5_fnames=[], n_subsample=100000,
		field_names=["Pressure (Pa)"],
		out_fname="cfd_pod_snapshots.h5"):
		out_fname="cfd_pod_snapshots.h5",
		vel_mult=1.0):
		"""
		Re-interpolates all data to be on a consistent mesh for POD
		modeling.
		@@ -248,6 +272,7 @@ def reinterpolate_cfd(

		# Summary data
		if "Velocity" in field_name:
		case_sub_field_data = vel_mult * case_sub_field_data
		summary_field_data = np.abs(case_sub_field_data)
		else:
		summary_field_data = case_sub_field_data
		@@ -287,6 +312,7 @@ def reinterpolate_cfd(
		if __name__ == "__main__":
		parser = argparse.ArgumentParser()
		parser.add_argument("-o", help="output h5 file name", type=str, default='cfd_pod_snapshots.h5')
		parser.add_argument("--vel_mult", help="Velocity vector multiplier", type=float, default=1.0)
		args = parser.parse_args()
		h5_fnames = []
		# field_names = [
		@@ -311,10 +337,13 @@ if __name__ == "__main__":
		print("Preprocessing file: %s" % csv_file)
		out_file = csv_file + ".h5"
		make_plot = False
		if i == 7:
		if i == 1:
		make_plot = True
		conv2hdf(run_dir + csv_file, outfile_name=out_file, case_params=params, make_plot=make_plot,
		csv_col_names=csv_col_names)
		h5_fnames.append(out_file)
		i += 1
		reinterpolate_cfd(h5_fnames[0], h5_fnames=h5_fnames, field_names=field_names, out_fname=args.o)
		base_id = max(0, int(len(h5_fnames) - len(h5_fnames) / 2))
		reinterpolate_cfd(h5_fnames[base_id], h5_fnames=h5_fnames,
		field_names=field_names, out_fname=args.o,
		vel_mult=args.vel_mult)

Admin message