Merge remote-tracking branch 'firemodels/master' into FireX

Author: cxp484

Manuals/FDS_User_Guide/FDS_User_Guide.tex

@@ -7043,7 +7043,9 @@ \subsection{Thermal Boundary Conditions at the Ground}
 
 \subsection{Example}
 
-Figure~\ref{ABL_profiles} displays velocity and temperature profiles generated by FDS over a 1000~m square domain with periodic boundaries and a height of 200~m. The wind fields are generated using pressure gradient forces, $F$, of various values, and the ground is given several different values of \ct{CONVECTIVE_HEAT_FLUX} ($\dot{q}_{\rm c}''$) and surface \ct{ROUGHNESS} ($s$). Eqs.~(\ref{eq:roughness_conversion}) and (\ref{qdot_L}) are used to convert the specified $\dot{q}_{\rm c}''$ and $s$ to $L$ and $z_0$ that are then used to generate Monin-Obukhov velocity and temperature profiles with which to compare the simulations. Note that these simulations do not invoke the Monin-Obukhov profiles directly. Rather, the M-O profiles are used to test if the FDS simulations produce realistic vertical profiles using just a specified pressure gradient force, $F$, surface roughness, $s$, and surface heat flux, $\dot{q}_{\rm c}''$.
+Figure~\ref{ABL_profiles} displays velocity and temperature profiles generated by FDS over a 1000~m square domain with periodic boundaries and a height of 400~m. The wind fields are generated using a pressure gradient force, $F$, of various values, and the ground is given several different values of \ct{CONVECTIVE_HEAT_FLUX} ($\dot{q}_{\rm c}''$) and surface \ct{ROUGHNESS} ($s$). Eqs.~(\ref{eq:roughness_conversion}) and (\ref{qdot_L}) are used to convert the specified $\dot{q}_{\rm c}''$ and $s$ to $L$ and $z_0$ that are then used to generate Monin-Obukhov velocity and temperature profiles with which to compare the simulations. Note that these simulations do not invoke the Monin-Obukhov profiles directly. Rather, the M-O profiles are used to test if the FDS simulations produce realistic vertical profiles from the specified pressure gradient force, $F$, surface roughness, $s$, and surface heat flux, $\dot{q}_{\rm c}''$. The simulations are initialized with a temperature lapse rate of -0.01~K/m, and the laterial boundary condition is \ct{'PERIODIC FLOW ONLY'}, meaning that the velocity field is periodic but the temperature field is not. The upper boundary is \ct{'OPEN'} with a specified \ct{SPEED} that is approximately the same as that which is predicted by M-O theory. In this regard, the profiles are not predicted based solely on the pressure gradient force and ground properties. Rather, lateral and ceiling conditions are needed to match the resulting profiles.
+
+The purpose of this example is to demonstrate that you can establish a realistic temperature and velocity profile even when the ground surface is not flat. Simulations like those shown here can be used to test if a given pressure gradient force and a ground roughness and heat flux are appropriate for a simulation over complex terrain.
 
 \begin{figure}[p]
    \begin{tabular*}{\textwidth}{l@{\extracolsep{\fill}}r}

Manuals/FDS_Validation_Guide/Experiment_Chapter.tex

@@ -620,7 +620,7 @@ \section{FHWA Tunnel}
 
 Reduced-scale tunnel experiments~\cite{FHWA:test_report} were conducted at the Institute f{\"u}r Angewandte Brandschutzforschung (IFAB), Germany, as part of a research program sponsored by the U.S. Federal Highway Administration (FHWA). The objective of the research was to assess operational integration of highway tunnel emergency ventilation systems with fixed fire fighting systems. The experiments assessed the ability of a sprinkler system to improve smoke management in a longitudinally vented tunnel. Measurements of velocity, temperature, pressure, humidity, and surface temperature were recorded, and video footage was taken. Three different nozzles were used in the tests and measurements of the nozzle spray patterns were reported.
 
-The tunnel was approximately 1/4 scale, with a width of 2.5~m, height of 1.25~m and a length of 12~m. The tunnel was lined with water-resistant and non-combustible cement board. Figure~\ref{FHWAtunnel1} provides a schematic diagram. 
+The tunnel was approximately 1/4 scale, with a width of 2.5~m, height of 1.25~m and a length of 12~m. The tunnel was lined with water-resistant and non-combustible cement board. Figure~\ref{FHWAtunnel1} provides a schematic diagram.
 
 Three fans were used to inject air into the tunnel and generate a longitudinal flow. To correlate fan operation to bulk longitudinal velocity, the inlet velocity profile was measured manually, in a cold flow test, at 25 points distributed over the inlet cross section, 1.5~m into the tunnel, using a handheld rotating vane anemometer. The bulk longitudinal velocity was approximately constant during the experiments, but the transient (fluctuating) profile was used in the simulations in order to better match the actual boundary conditions.
 
@@ -630,11 +630,11 @@ \section{FHWA Tunnel}
 
 \subsubsection{Modeling Notes}
 
-A summary of the experiments is provided in Table~\ref{tab:FHWA_Test_Matrix}. The FDS simulations used a cubic grid resolution of 5~cm throughout. Different spray nozzles were tested to measure the interaction between the water spray and the ventilation. Typically, the nozzles were activated after some back-layering had been visually observed. 
+A summary of the experiments is provided in Table~\ref{tab:FHWA_Test_Matrix}. The FDS simulations used a cubic grid resolution of 5~cm throughout. Different spray nozzles were tested to measure the interaction between the water spray and the ventilation. Typically, the nozzles were activated after some back-layering had been visually observed.
 
 
 \begin{figure}[!ht]
-\includegraphics[width=\textwidth]{FIGURES/FHWA_Tunnel/Nozzle_A.jpg} 
+\includegraphics[width=\textwidth]{FIGURES/FHWA_Tunnel/Nozzle_A.jpg}
 \includegraphics[width=\textwidth]{FIGURES/FHWA_Tunnel/Nozzle_B.jpg}
 \caption[Sketches of the FHWA/IFAB experiments]{Sketches of the two sprinkler configurations in the FHWA/IFAB experiments. The fans and flow straightener are to the left. The cross-hatched squares are access panels. The burner is below Nozzle~A.}
 \label{FHWAtunnel1}
@@ -1088,7 +1088,7 @@ \section{Lattimer Corridor Ceiling}
 
 Lattimer et al. studied the thermal environment created by a fire impinging on a ceiling at the end of a corridor in~\cite{Lattimer:FTJ:2013}. The apparatus, shown in Figure~\ref{fig:lattimer}, consisted of a 2.4~m long 1.2~m wide corridor with a ceiling height of 2.1~m from the floor. The back wall and back 1.2~m of the side walls extended 1.2~m below the ceiling. The remaining 1.2~m of the side walls were extended 0.6~m below the ceiling. The overall apparatus was elevated 0.9~m off of the floor to allow air to flow into the bottom of the corridor from all sides. A 0.46~m deep by 1.15~m wide propane sand burner was centered on the back wall with the top surface located either 0.6~m or 1.1~m from the ceiling. Each separation distance was tested at four heat release rates ranging from 100-400~kW.
 
-The authors measured the water cooled gauge heat flux and gas temperature at four locations along the ceiling. The distances from the back wall were 0.3~m, 0.9~m, 1.5~m, and 2.1~m. Table~\ref{tab:lattimer_ceiling_exp} summarizes the test conditions and measurements from this test series. 
+The authors measured the water cooled gauge heat flux and gas temperature at four locations along the ceiling. The distances from the back wall were 0.3~m, 0.9~m, 1.5~m, and 2.1~m. Table~\ref{tab:lattimer_ceiling_exp} summarizes the test conditions and measurements from this test series.
 
 \begin{table*}[!ht]
 \centering
@@ -1164,7 +1164,7 @@ \section{LEMTA Spray Test for Radiation Attenuation}
 \section{LEMTA Spray Cooling}
 \label{LEMTA_Spray_Cooling_Description}
 
-A series of experiments was conducted at the Laboratoire \'Energies and M\'ecanique Th\'eorique et Appliqu\'ee (LEMTA) to measure the temperature of a hot steel plate cooled by a water spray~\cite{Acem:ISFEH2022}. The 1~m by 1~m by 2~mm thick steel plate, in either a horizontal or vertical configuration, was heated by a radiant panel either 20~cm below or to the side. The 50~cm by 50~cm heater delivered a total power of approximately 50~kW with a radiative component estimated to be 25~kW, producing a radiative heat flux near 100~kW/m$^2$.  After the plate reached a steady temperature, the radiant panel was switched off and a water spray nozzle 50~cm above or to the side of the plate was activated. Three different nozzle designs were tested: (1) a single Protectospray D3 (PSD3) conical nozzle from Tyco flowing at 34.5~L/min, (2) a single SU42 conical jet nozzle from Spraying Systems flowing at 4.6~L/min, and (3) four TPU400067 flat jet nozzles from Spraying Systems flowing for a total of 1.6~L/min. 
+A series of experiments was conducted at the Laboratoire \'Energies and M\'ecanique Th\'eorique et Appliqu\'ee (LEMTA) to measure the temperature of a hot steel plate cooled by a water spray~\cite{Acem:ISFEH2022}. The 1~m by 1~m by 2~mm thick steel plate, in either a horizontal or vertical configuration, was heated by a radiant panel either 20~cm below or to the side. The 50~cm by 50~cm heater delivered a total power of approximately 50~kW with a radiative component estimated to be 25~kW, producing a radiative heat flux near 100~kW/m$^2$.  After the plate reached a steady temperature, the radiant panel was switched off and a water spray nozzle 50~cm above or to the side of the plate was activated. Three different nozzle designs were tested: (1) a single Protectospray D3 (PSD3) conical nozzle from Tyco flowing at 34.5~L/min, (2) a single SU42 conical jet nozzle from Spraying Systems flowing at 4.6~L/min, and (3) four TPU400067 flat jet nozzles from Spraying Systems flowing for a total of 1.6~L/min.
 
 Each experiment was performed in a horizontal and vertical orientation. The steel plate temperature was measured at the center on top and bottom in order to assess the heat transfer across the plate.
 
@@ -2700,7 +2700,7 @@ \subsubsection{Modeling Notes}
 Materials for which data was not available in the original studies used a $Y_{s}$ of 0.05 g/g.
 
 Note that the RISE and FPL databases did not provide ignition properties or information on specific material conditioning or modifications to test procedures.
-For these materials, 
+For these materials,
 a thermal conductivity of 0.4 $\mathrm{W/(m\cdot K)}$,
 specific heat capacity of 1.0 $\mathrm{kJ/(kg\cdot K)}$,
 and an emissivity of unity was assumed.
@@ -3018,13 +3018,13 @@ \section{TUS Facade Experiments}
 
 The compartment dimension were 4.0~m by 4.0~m by 1.7~m. Note that the dimensions are incorrect in Ref.~\cite{Yoshioka:FST2012}. The facade was 5.7~m tall, with a 1.2~m wide side wall made of the same material as the main facade. The fire compartment had one 1.2~m tall and 2.0~m wide window in the facade wall, but the window was partially closed in two of the experiments. The rear of the fire compartment had four 0.5~m by 0.6~m openings that were open or closed according to the desired ventilation condition. These openings were closed by heavy fire blankets, and there was considerable uncertainty in the actual opening area.
 
-A mixture of gaseous fuels supplied by the local utility was used as a fuel. It was supplied to the compartment through four, room-wide burner pipes with holes on the top surface, approximately 20~cm above the floor. The experiments were conducted in three series, named I, 1 and 2. For each, multiple fires were generated between which changes were made to the ventilation and HRR. The test duration was 5~min with 10~min of cooling between each. 
+A mixture of gaseous fuels supplied by the local utility was used as a fuel. It was supplied to the compartment through four, room-wide burner pipes with holes on the top surface, approximately 20~cm above the floor. The experiments were conducted in three series, named I, 1 and 2. For each, multiple fires were generated between which changes were made to the ventilation and HRR. The test duration was 5~min with 10~min of cooling between each.
 
 A second set of experiments was performed in this facility by Sun et~al.~\cite{Sun:FAM2024} using a configuration based on the JIS~A~1310 test standard. The fire compartment for this test is 1.35~m by 1.35~m by 1.35~m with a 0.91~m by 0.91~m opening. The 0.6~m by 0.6~m propane burner is flush with the floor at the rear of the compartment with heat release rates of 600~kW, 750~kW, and 900~kW.
 
 \subsubsection{Modeling Notes}
 
-Gleb Bytskov at the University of Aalto, Finland, performed numerical simulations of these experiments using FDS as part of his master's thesis~\cite{Bytskov:Thesis}. The grid resolution is 10~cm. The fuel is assumed to be propane with a soot yield of 0.015, CO yield of 0.015, and radiative fraction of 0.20. 
+Gleb Bytskov at the University of Aalto, Finland, performed numerical simulations of these experiments using FDS as part of his master's thesis~\cite{Bytskov:Thesis}. The grid resolution is 10~cm. The fuel is assumed to be propane with a soot yield of 0.015, CO yield of 0.015, and radiative fraction of 0.20.
 
 The JIS~A~1310 simulations use propane with a radiative fraction of 0.30.
 
@@ -3618,19 +3618,19 @@ \section{UWO Wind Tunnel Experiments}
 
 \subsubsection{Modeling Notes}
 
-The 1:100 scale model of the building is approximately 14~cm long, 9~cm wide, and 4~cm tall. The simulations use 5.0~mm, 2.5~mm, and 1.25~mm cubic grid cells, uniformly spanning the computational domain of dimensions 1.12~m $\times$ 0.56~m $\times$ 0.28~m. The 4~cm height of the scale model is spanned by 8, 16, and 32 grid cells. The grid resolution was assessed by checking that the values of $y^+$, a non-dimensional distance, never exceed 150 for any mesh resolution. $y^+$ values are dependent on turbulence model wall laws and a value of 150 means that the first grid point is well within the logarithmic region.
+The 1:100 scale model of the Test~7 building is approximately 14~cm long, 9~cm wide, and 4~cm tall. The simulations use 5.0~mm, 2.5~mm, and 1.25~mm cubic grid cells, uniformly spanning the computational domain of dimensions 1.12~m $\times$ 0.56~m $\times$ 0.28~m. The 4~cm height of the scale model is spanned by 8, 16, and 32 grid cells. The grid resolution was assessed by checking that the values of $y^+$, a non-dimensional distance, never exceed 150 for any mesh resolution. $y^+$ values are dependent on turbulence model wall laws and a value of 150 means that the first grid point is well within the logarithmic region.
 
 Only two wind directions are chosen for validation, 180\si{\degree} and 270\si{\degree}, representing right angles formed with the long and short sides of the building. For each angle, 3 lines of pressure measurements running up the windward side, along the roof, and down the leeward side are used for comparison to the simulation. One line of pressure measurements along the side of the scale model is also used.
 
-The incoming flow assigned at the inflow boundary is characterized by its mean wind speed profile and turbulent fluctuations. For consistency with the UWO experiments, the mean wind speed profile was modeled by the power law:
+The incoming flow is defined using an \ct{OPEN} wind boundary condition, and is characterized by its mean wind speed profile and turbulent fluctuations. For consistency with the UWO experiments, the mean wind speed profile was modeled by the power law:
 \begin{equation}
 U(z) = 9.144 \left(\frac{z}{0.0396}\right)^p
 \end{equation}
-where 9.144 m/s is the reference wind speed at the 0.0396~m roof height of the scale model and the power law exponent $p=0.1173$ was fit to the experimental data.
+where 9.144 m/s is the reference wind speed at the 0.0396~m roof height of the scale model and the power law exponent $p=0.1173$ was fit to the experimental data. The vertical wind profile $U(z)$ is applied using a \ct{RAMP} function.
 
-Fluctuations at the inlet boundary are generated by the Synthetic Eddy Method (SEM) of Jarrin~\cite{Jarrin:2008}. The desired value of \ct{N_EDDY} (the number of eddies generated at the inlet) is the largest possible value that does not slow down the simulation. This value is determined by running a number of simulations for the same time while varying the \ct{N_EDDY} input and recording the total run time. The other turbulence parameters, \ct{L_EDDY} (characteristic eddy length) and \ct{REYNOLDS_STRESS}, are set based on the turbulence intensity data from the UWO experiments. The Reynolds stresses are found by multiplying the turbulence intensity in each direction by the wind velocity at roof height to determine the root mean square of the velocity fluctuations, and then squaring the root mean square values.
+Fluctuations at the inflow boundary are generated by the Synthetic Eddy Method (SEM) of Jarrin~\cite{Jarrin:2008}. The desired value of \ct{N_EDDY} (the number of eddies generated at the inlet) is the largest possible value that does not slow down the simulation. This value is determined by running a number of simulations for the same time while varying the \ct{N_EDDY} input and recording the total run time. The other turbulence parameters, \ct{L_EDDY} (characteristic eddy length) and \ct{REYNOLDS_STRESS}, are set based on the turbulence intensity data from the UWO experiments. The Reynolds stresses are found by multiplying the turbulence intensity in each direction by the wind velocity at roof height to determine the root mean square of the velocity fluctuations, and then squaring the root mean square values.
 
-The side walls and roof of the wind tunnel are set to free-slip. The roughness length at the ground is specified to be 0.0001~m, 1/100 the full-scale value of 0.01~m. The outlet of the wind tunnel is set to an ``open'' boundary condition, where pressure is set to the ambient pressure.
+The side walls and roof of the wind tunnel are set to free-slip. The roughness length at the ground is specified to be 0.0001~m, 1/100 the full-scale value of 0.01~m. The outlet of the wind tunnel is also set to an \ct{OPEN} boundary condition, where pressure is set to the ambient pressure.
 
 
 \section{Vettori Flat Ceiling Experiments}