EXP: add heat transfer coefficient data from Bouchair thesis

Bouchair_Solar_Chimney/Bouchair_Heat_Transfer_Coefficients.csv

@@ -0,0 +1,25 @@
+case,inlet_height_m,cavity_gap_m,wall_temperature_C,mdot_exp_kg_s_m,mdot_exp_kg_s,htc_exp_W_m2_K,thesis_table
+SC_p1_p1_30,0.1,0.1,30,0.049,0.0735,6.7,4.29
+SC_p1_p1_40,0.1,0.1,40,0.065,0.0975,8.2,4.29
+SC_p1_p1_50,0.1,0.1,50,0.080,0.1200,8.9,4.29
+SC_p1_p1_60,0.1,0.1,60,0.090,0.1350,9.8,4.29
+SC_p1_p2_30,0.1,0.2,30,0.060,0.0900,7.0,4.29
+SC_p1_p2_40,0.1,0.2,40,0.082,0.1230,7.9,4.29
+SC_p1_p2_50,0.1,0.2,50,0.099,0.1485,9.3,4.29
+SC_p1_p2_60,0.1,0.2,60,0.110,0.1650,9.7,4.29
+SC_p1_p3_30,0.1,0.3,30,0.059,0.0885,5.6,4.29
+SC_p1_p3_40,0.1,0.3,40,0.080,0.1200,6.5,4.29
+SC_p1_p3_50,0.1,0.3,50,0.095,0.1425,7.4,4.29
+SC_p1_p3_60,0.1,0.3,60,0.109,0.1635,7.8,4.29
+SC_p4_p1_30,0.4,0.1,30,0.046,0.0690,4.9,4.30
+SC_p4_p1_40,0.4,0.1,40,0.067,0.1005,5.9,4.30
+SC_p4_p1_50,0.4,0.1,50,0.080,0.1200,6.6,4.30
+SC_p4_p1_60,0.4,0.1,60,0.086,0.1290,6.8,4.30
+SC_p4_p2_30,0.4,0.2,30,0.064,0.0960,6.1,4.30
+SC_p4_p2_40,0.4,0.2,40,0.094,0.1410,6.7,4.30
+SC_p4_p2_50,0.4,0.2,50,0.103,0.1545,7.0,4.30
+SC_p4_p2_60,0.4,0.2,60,0.118,0.1770,7.5,4.30
+SC_p4_p3_30,0.4,0.3,30,0.053,0.0795,4.5,4.30
+SC_p4_p3_40,0.4,0.3,40,0.097,0.1455,6.1,4.30
+SC_p4_p3_50,0.4,0.3,50,0.107,0.1605,6.5,4.30
+SC_p4_p3_60,0.4,0.3,60,0.121,0.1815,6.7,4.30

Bouchair_Solar_Chimney/README.md

@@ -0,0 +1,47 @@
+# Global convective heat transfer coefficients
+
+Source: A. Bouchair, thesis, `Bouchair_THESIS_uk_bl_ethos_329792.pdf`,
+Tables 4.29 and 4.30, printed page 104 (PDF page 113). The CSV transcribes
+all 24 HTC values and their accompanying mass flow rates as printed.
+HTC units are W/(m2 K). `mdot_exp_kg_s_m` is mass flow per unit cavity
+span in kg/(s m). `mdot_exp_kg_s` is the total mass flow in kg/s, calculated
+as `mdot_exp_kg_s_m * 1.5 m`.
+
+The four columns correspond to wall temperatures 10, 20, 30, and 40 K above
+20 C ambient, mapped here to FDS case temperatures 30, 40, 50, and 60 C.
+See also Tables 4.25 and 4.26 (printed page 99) and Tables 4.33 and 4.34
+(printed page 107). The mass flows are retained from the HTC tables:
+Table 4.29 gives 0.095 for the 0.1 m inlet, 0.3 m gap, 50 C case, whereas
+Table 4.25 gives 0.097; Table 4.30 gives 0.053 for the 0.4 m inlet, 0.3 m
+gap, 30 C case, whereas Table 4.26 gives 0.060. No corrections are applied.
+
+Equation 4.3, printed page 103 (PDF page 112), defines
+
+    hc = M cp (T_out - T_in) / [2 Y Z (T_wall - T_bulk)]
+    T_bulk = (T_in + T_out)/2
+
+For total mass flow M in kg/s, use cavity span Y=1.5 m and height Z=2.0 m,
+giving a nominal two-wall area of 6 m2. Equivalently, flow per unit span
+uses area per unit span 2Z=4 m. The companion FDS script uses total `mdot`,
+cp=1000 J/(kg K), T_in=20 C, and separate temporal means of `mdot` and
+the spatially averaged `T_outlet` over the last half of the available run.
+The area follows Eq. 4.3 without deducting the cavity inlet opening.
+
+Bouchair excludes 0.5 and 1.0 m gaps from these HTC tables because of
+flow complexity (printed page 103); no experimental HTC is assigned to them.
+
+## Python comparison plot
+
+`Utilities/Python/scripts/bouchair_htc.py` in the FDS repository writes
+`Bouchair_HTC.csv` and `Bouchair_HTC.pdf`. The PDF plots positive matched
+experimental/predicted HTC pairs on equal log scales. Black lines show
+unity and experimental spread; red lines show model bias and model spread.
+The statistics use the ten-bin weighting and log-ratio formulas from
+`fdsplotlib.scatplot`. Dashed bounds use factors `1+2*sigma`, as in FDS.
+
+Supply `--sigma-e` as a relative standard deviation (fraction) when an HTC
+experimental uncertainty is available. If omitted, the FDS fallback sets
+experimental and model variances equal to half the log-ratio variance;
+the PDF explicitly labels the experimental uncertainty as estimated.
+This fallback is not a thesis-reported uncertainty. The 0.08 mass-flow
+uncertainty from the reference figure is not assigned to HTC.