Skip to content

Commit

Permalink
Merge pull request #362 from rmcdermo/master
Browse files Browse the repository at this point in the history
Liquid_Pool_Fires: reorganize NIST cmp config and readme
  • Loading branch information
rmcdermo committed Aug 18, 2023
2 parents 5c15946 + ea6ca8d commit d47f1eb
Show file tree
Hide file tree
Showing 95 changed files with 212 additions and 21 deletions.
4 changes: 3 additions & 1 deletion .github/workflows/Liquid_Pool_Fires.yml
Original file line number Diff line number Diff line change
Expand Up @@ -53,7 +53,8 @@ jobs:
- name: Push NIST NIST Pool Fires Results to release
if: github.event_name == 'push' && github.repository_owner == 'MaCFP'
run: |
cd $GITHUB_WORKSPACE/Liquid_Pool_Fires/NIST_Pool_Fires/Computational_Results/2023/NIST/Preliminary_Results
cd $GITHUB_WORKSPACE/Liquid_Pool_Fires/NIST_Pool_Fires/Computational_Results/2023/NIST
cp README.md ./Plots/.
zip Plots.zip -r Plots
mv Plots.zip NIST_Pool_Fires_NIST_Plots.zip
gh release upload SCRIPT_FIGURES NIST_Pool_Fires_NIST_Plots.zip --clobber
Expand All @@ -72,6 +73,7 @@ jobs:
if: github.event_name == 'push' && github.repository_owner == 'MaCFP'
run: |
cd $GITHUB_WORKSPACE/Liquid_Pool_Fires/Waterloo_Methanol/Computational_Results/2023/NIST
cp README.md ./Plots/.
zip Plots.zip -r Plots
mv Plots.zip Waterloo_Methanol_NIST_Plots.zip
gh release upload SCRIPT_FIGURES Waterloo_Methanol_NIST_Plots.zip --clobber
Expand Down

Large diffs are not rendered by default.

Original file line number Diff line number Diff line change
Expand Up @@ -19,8 +19,8 @@
institute='NIST',
revision='MaCFP-3, Tsukuba, 2023',
expdir='../../../Experimental_Data/',
cmpdir='./Preliminary_Results/',
pltdir='./Preliminary_Results/Plots/',
cmpdir='./Output/',
pltdir='./Plots/',
close_figs=True,
verbose=True,
plot_list=['all'])
Expand Down
Original file line number Diff line number Diff line change
@@ -0,0 +1,207 @@

### Contributor
Name: Randy McDermott<sup>a</sup>, Kevin McGrattan<sup>a</sup>, Jason Floyd<sup>b</sup>

Institutions:
<sup>a</sup>National Institute of Standards and Technology (NIST), Gaithersburg, Maryland
<sup>b</sup>Fire Safety Research Institute, UL Research Institutes, Columbia, Maryland

Country: USA

------------------

### Test case

Case 3a: Waterloo Methanol 30 cm

------------------

### CFD package
Code: [Fire Dynamics Simulator (FDS)](https://github.com/firemodels/fds)

Version:

FDS-6.8.0-564-g879930a-master

------------------

### Resolution

#### Computational domain discretization (flow solver)
Domain: 0.6 m x 0.6 m by 1.2 m

Cell size: 5 mm, 1 cm, 2 cm

Cell type: cubic

Total cells: 5 mm case: 2304000; 1 cm case: 288000; 2 cm case: 36000

Comments:

#### Angular space discretization (radiation solver)
Number of solid angles: 600, `TIME_STEP_INCREMENT`=10, `ANGLE_INCREMENT`=10

Comments: Default FDS radiation solver

------------------

### Initial conditions
Comments: All flow variables are originally ambient and then ramped up in approximately 1 s

------------------

### Boundary conditions
* OPEN at all exterior boundaries
* log law wall function for momentum
* combined forced and natural convection correlations for h_conv and h_mass

#### Predicted MLR (methanol)
* 10 cm thick SURF with 15 cells in depth
* ∆z = 0.2618 mm in first liquid cell based on ∆z<sup>2</sup>/α<sup>2</sup> = 1 (unity time scale))
* stretched grid with stretch factor of 2
* liquid properties for methanol; no effective turblent transport
* in-depth radiation absorption considered; κ=1140 m<sup>-1</sup>
* dynamic, energy-conserving remeshing scheme as liquid thickness recedes

Comments:

------------------

### Models (include parameters)
Turbulence model: Deardorff (algebraic k_sgs, C_DEARDORFF=0.1), Sc_t = Pr_t = 0.5

Near-wall turb model: WALE (C_WALE=0.6)

Combustion model: EDM wih two-step fast-fast serial reactions, see table below for fuel-specific parameters, no suppression

Radiation model: Finite-volume, upwind, first-order accurate

Radiative fraction: See table

| Fuel | Radiant Fraction | FUEL_C_TO_CO_FRACTION | FUEL_H_TO_H2_FRACTION |
| :-----------------| :----------------| :---------------------| :---------------------|
| Acetone | 0.31 | 0.85 | 0.50 |
| Ethanol | 0.26 | 0.95 | 0.50 |
| Methane | 0.15 | 0.97 | 0.50 |
| Methanol (30 cm) | 0.22 | 0.97<sup>1</sup> | 0.50 |
| Methanol (100 cm) | 0.21 | 0.97<sup>1</sup> | 0.50 |
| Propane (20 kW) | 0.15 | 0.85 | 0.50 |
| Propane (34 kW) | 0.22 | 0.85 | 0.50 |

Comments: A two-step reaction mechanism is implemented. In the first reaction, fuel is converted to CO, soot, H2, and H2O. In the second reaction, the CO, soot, and H2 are converted to CO2 and H2O. Both reactions employ fast kinetics, but proceed in series, not in parallel. The relative amounts of CO, soot, and H2 produced in the first step (dictated by the parameters FUEL_C_TO_CO_FRACTION and FUEL_H_TO_H2_FRACTION) are still subjects of study, and for the moment have been estimated based on measured results. The fractions of carbon atoms converted to CO in the first step are as follows---0.85 for acetone; 0.95 for ethanol; 0.97 for methane; 0.97 for methanol<sup>1</sup> ; 0.85 for propane. For all fuels, one half of the hydrogen atoms are converted to H2 in the first step.

<sup>1</sup>The cases run for MaCFP-2 used FUEL_C_TO_CO_FRACTION=0.97 for methanol. But since then we have changed the input files in our repository to reflect no in-flame soot formation, i.e., FUEL_C_TO_CO_FRACTION=1.

------------------

### Discretization methods
Time: Predictor-Corrector; second-order accurate

CFL: 0.8 to 1.0

Advection: CHARM flux limiter

Diffusion: Second-order accurate central difference

Pressure-velocity coupling: Low Mach number approximation; solution of Poisson equation for pressure

------------------

### Computational Cost (hh:mm:ss)

#### Acetone

| ∆x (cm) | Total cells | CPU cores | Sim Time (s) | Wall Clock (s) | WC (dd-hh:mm:ss) | CPU Cost |
| :-------| :-----------| :-------- | :----------- | :------------- | :--------------- | :------- |
| 2 | 54000 | 12 | 60 | 2927 | 00-00:48:47 | 0.0108 |
| 1 | 432000 | 12 | 60 | 35481 | 00-09:51:21 | 0.0164 |
| 0.5 | 3456000 | 24 | 60 | 231898 | 02-16:24:58 | 0.0268 |

#### Ethanol

| ∆x (cm) | Total cells | CPU cores | Sim Time (s) | Wall Clock (s) | WC (dd-hh:mm:ss) | CPU Cost |
| :-------| :-----------| :-------- | :----------- | :------------- | :--------------- | :------- |
| 2 | 54000 | 12 | 60 | 2895 | 00-00:48:47 | 0.0107 |
| 1 | 432000 | 12 | 60 | 31254 | 00-09:51:21 | 0.0144 |
| 0.5 | 3456000 | 24 | 60 | 223101 | 02-16:24:58 | 0.0258 |

#### Methane

| ∆x (cm) | Total cells | CPU cores | Sim Time (s) | Wall Clock (s) | WC (dd-hh:mm:ss) | CPU Cost |
| :-------| :-----------| :-------- | :----------- | :------------- | :--------------- | :------- |
| 2 | 54000 | 24 | 60 | 1858 | 00-00:30:58 | 0.0137 |
| 1 | 432000 | 24 | 60 | 20214 | 00-05:36:54 | 0.0187 |
| 0.5 | 3456000 | 24 | 60 | 219842 | 02-13:04:02 | 0.0254 |

#### Methanol 30 cm prescribed MLR

600 radiation angles for predicting heat flux

| ∆x (cm) | Total cells | CPU cores | Sim Time (s) | Wall Clock (s) | WC (dd-hh:mm:ss) | CPU Cost |
| :-------| :-----------| :-------- | :----------- | :------------- | :--------------- | :------- |
| 2 | 72000 | 16 | 60 | 2587 | 00-00:43:07 | 0.0095 |
| 1 | 576000 | 16 | 60 | 31881 | 00-08:51:21 | 0.0147 |
| 0.5 | 4608000 | 32 | 60 | 229951 | 02-15:52:31 | 0.0266 |

#### Methanol 100 cm prescribed MLR

500 radiation angles for predicting heat flux

| ∆x (cm) | Total cells | CPU cores | Sim Time (s) | Wall Clock (s) | WC (dd-hh:mm:ss) | CPU Cost |
| :-------| :-----------| :-------- | :----------- | :------------- | :--------------- | :------- |
| 4 | 270000 | 10 | 60 | 7925 | 00-02:12:05 | 0.0048 |
| 2 | 648000 | 24 | 60 | 22562 | 00-06:16:02 | 0.0139 |
| 1 | 3744000 | 108 | 60 | 64430 | 00-17:53:50 | 0.0309 |

#### Methanol 100 cm predicted MLR

500 radiation angles for predicting heat flux

| ∆x (cm) | Total cells | CPU cores | Sim Time (s) | Wall Clock (s) | WC (dd-hh:mm:ss) | CPU Cost |
| :-------| :-----------| :-------- | :----------- | :------------- | :--------------- | :------- |
| 4 | 270000 | 10 | 60 | 7218 | 00-02:00:18 | 0.0044 |
| 2 | 648000 | 24 | 60 | 30640 | 00-08:30:40 | 0.0189 |
| 1 | 3744000 | 108 | 60 | 67548 | 00-18:45:48 | 0.0324 |

#### Propane 20 kW

| ∆x (cm) | Total cells | CPU cores | Sim Time (s) | Wall Clock (s) | WC (dd-hh:mm:ss) | CPU Cost |
| :-------| :-----------| :-------- | :----------- | :------------- | :--------------- | :------- |
| 2 | 54000 | 24 | 60 | 1558 | 00-00:25:58 | 0.0115 |
| 1 | 432000 | 24 | 60 | 18227 | 00-05:03:47 | 0.0168 |
| 0.5 | 3456000 | 24 | 60 | 163774 | 01-21:29:34 | 0.0189 |

#### Propane 34 kW

| ∆x (cm) | Total cells | CPU cores | Sim Time (s) | Wall Clock (s) | WC (dd-hh:mm:ss) | CPU Cost |
| :-------| :-----------| :-------- | :----------- | :------------- | :--------------- | :------- |
| 2 | 54000 | 24 | 60 | 1555 | 00-00:25:55 | 0.0115 |
| 1 | 432000 | 24 | 60 | 17320 | 00-04:48:40 | 0.0160 |
| 0.5 | 3456000 | 24 | 60 | 190057 | 02-04:47:37 | 0.0220 |

Wall clock time: see table above

Simulation time: see table above

Number of CPUs (MPI Processes): see table above

CPU cost (Number of CPUs * Wall clock time / Simulation time / Total cells): see table above

------------------

### Averaging period

total simulation time: 60 s
averaged statistics: last 50 s

------------------

### Special issues/problems

------------------

### Relevant publications
1. [FDS Technical Reference Guide](https://github.com/firemodels/fds/releases)

2. [FDS Validation Guide](https://github.com/firemodels/fds/releases)

0 comments on commit d47f1eb

Please sign in to comment.