Merge pull request #15196 from rmcdermo/master

Python: add umd_line_burner_process.py

Author: rmcdermo

Utilities/Python/scripts/umd_line_burner_process.py

@@ -0,0 +1,74 @@
+#!/usr/bin/env python3
+# McGrattan
+# 2-26-2018
+# umd_line_burner_process.py
+#
+# Read and process FDS output files for UMD Line Burner
+
+import os
+import numpy as np
+import pandas as pd
+
+Lf_dt = 10
+
+outdir = '../../../out/UMD_Line_Burner/'
+
+fuel_name = ['methane', 'propane']
+res_name = ['1p25cm', 'p625cm', 'p3125cm']
+
+for i_fuel in range(2):
+
+    for fds_resolution in range(3):
+
+        dev_file = os.path.join(outdir, f"{fuel_name[i_fuel]}_dx_{res_name[fds_resolution]}_devc.csv")
+        hrr_file = os.path.join(outdir, f"{fuel_name[i_fuel]}_dx_{res_name[fds_resolution]}_hrr.csv")
+
+        # Read CSVs, skipping first two header lines
+        DEV = pd.read_csv(dev_file, skiprows=2, header=None)
+        HRR = pd.read_csv(hrr_file, skiprows=2, header=None)
+
+        # Read headers separately (the first row after skiprows=1)
+        with open(dev_file, 'r') as f:
+            header_lines = [next(f) for _ in range(2)]
+        dev_headers = header_lines[1].strip().split(',')
+
+        with open(hrr_file, 'r') as f:
+            header_lines = [next(f) for _ in range(2)]
+        hrr_headers = header_lines[1].strip().split(',')
+
+        # Locate columns (match MATLAB’s strcmp behavior exactly)
+        Time_idx = dev_headers.index('Time')
+        XO2_idx = dev_headers.index('"XO2"')
+        qrad1_idx = dev_headers.index('"qrad1"')
+        qrad2_idx = dev_headers.index('"qrad2"')
+        Lf_idx = dev_headers.index('"L_F"')
+
+        HRR_idx = hrr_headers.index('HRR')
+        Qrad_idx = hrr_headers.index('Q_RADI')
+
+        Time_FDS = DEV.iloc[:, Time_idx].values
+        XO2_FDS = DEV.iloc[:, XO2_idx].values
+        Qdot_FDS = HRR.iloc[:, HRR_idx].values
+        Qrad_FDS = HRR.iloc[:, Qrad_idx].values
+        q_R_FDS = 0.5 * (DEV.iloc[:, qrad1_idx].values + DEV.iloc[:, qrad2_idx].values)
+        Lf_FDS = DEV.iloc[:, Lf_idx].values.copy()
+
+        ntp = len(Time_FDS)
+
+        # Moving-average Lf_FDS with Lf_dt window
+        Lf_tmp = Lf_FDS.copy()
+        for n in range(ntp):
+            mask = Time_FDS > (Time_FDS[n] - Lf_dt)
+            # indices from first True to n inclusive
+            idxs = np.where(mask)[0]
+            idxs = idxs[idxs <= n]
+            Lf_FDS[n] = np.mean(Lf_tmp[idxs])
+
+        # Write output file
+        out_file = os.path.join(outdir, f"{fuel_name[i_fuel]}_dx_{res_name[fds_resolution]}.csv")
+        with open(out_file, 'w') as fid:
+            fid.write('XO2,eta,Chi_R,Lf,q_R\n')
+            for ii in range(ntp):
+                eta = Qdot_FDS[ii] / 50.0
+                Chi_R = max(0, -Qrad_FDS[ii] / max(0.001, Qdot_FDS[ii]))
+                fid.write(f"{XO2_FDS[ii]:5.3f},{eta:6.2f},{Chi_R:6.2f},{Lf_FDS[ii]:6.2f},{q_R_FDS[ii]:6.2f}\n")