ISSUE WITH DROPLET INSERTION THROUGH CURVED SURFACE CREATED USING GEOM

[mohanvadaga]

I am trying to inject water droplets at 285m/s into still air through a curved surface. I have created the surface in blender and imported it to fds. However the droplet velocity is suddenly reducing to near 0 immediately after insertion. I have also tried using the default sphere geom of fds still facing the same problem.

[marcosvanella]

@mohanvadaga post a simple case that demostrates this problem.

Also, 285 m/s is Mach~0.83 in atmospheric conditions, way beyond the aplicability of FDS.

[drjfloyd]

From the Technical Reference Guide:

[mohanvadaga]

Dear Marco and floyd, agreed, but I am injecting only droplets at such speed. The fluid flow is however governed by drag which I am temporarily reducing using primary breakup length and primary breakup drag reduction factor. I would like to know whether fds supports SURF IDs with VEL_PART on GEOM IDs. Because first I tried a simple case with obstruction where one of the surface is assigned with injection using VEL=0 and Vel_part=-285,and it worked fine. When I changed the obstruction to GEOM the velocity of particles dropped to zero however when I changed velocity to some value like vel=-50, the particles also started moving but after some time may be due to drag.

[mohanvadaga]

My code:
&HEAD CHID="Test-1"/
&TIME T_END=5, DT=0.001/
&DUMP DT_RESTART=2, NFRAMES=1000,DT_DEVC=0.001,VELOCITY_ERROR_FILE=.TRUE.,MAXIMUM_PARTICLES=1000000/
&MISC SIMULATION_MODE='LES', TURBULENCE_MODEL='DEARDORFF',POROUS_FLOOR=F,CFL_MAX=0.1, PARTICLE_CFL_MAX=0.1,PARTICLE_CFL=.TRUE./
&PRES SUSPEND_PRESSURE_ITERATIONS=.TRUE.,MAX_PRESSURE_ITERATIONS=50,PRESSURE_TOLERANCE=0.000001,VELOCITY_TOLERANCE=0.000001,SOLVER='UGLMAT'/

&MESH ID='Mesh-01', IJK=25,40,45, XB=0.0,25.0,20.0,60.0,45.0,90.0, MPI_PROCESS=0/

&MESH ID='Mesh-02', IJK=25,40,45, XB=25.0,50.0,-20.0,20.0,0.0,45.0, MPI_PROCESS=1/
&MESH ID='Mesh-03', IJK=25,40,45, XB=25.0,50.0,-20.0,20.0,45.0,90.0, MPI_PROCESS=2/
&MESH ID='Mesh-04', IJK=25,40,45, XB=25.0,50.0,-60.0,-20.0,0.0,45.0, MPI_PROCESS=3/
&MESH ID='Mesh-05', IJK=25,40,45, XB=25.0,50.0,-60.0,-20.0,45.0,90.0, MPI_PROCESS=4/
&MESH ID='Mesh-06', IJK=25,40,45, XB=25.0,50.0,20.0,60.0,0.0,45.0, MPI_PROCESS=5/
&MESH ID='Mesh-07', IJK=25,40,45, XB=25.0,50.0,20.0,60.0,45.0,90.0, MPI_PROCESS=6/
&MESH ID='Mesh-08', IJK=25,40,45, XB=0.0,25.0,-20.0,20.0,45.0,90.0, MPI_PROCESS=7/

&MESH ID='Mesh-09', IJK=25,40,45, XB=0.0,25.0,-60.0,-20.0,0.0,45.0, MPI_PROCESS=8/
&MESH ID='Mesh-10', IJK=25,40,45, XB=0.0,25.0,-60.0,-20.0,45.0,90.0, MPI_PROCESS=9/
&MESH ID='Mesh-11', IJK=25,40,45, XB=0.0,25.0,20.0,60.0,0.0,45.0, MPI_PROCESS=10/
&MESH ID='Mesh-12', IJK=25,40,45, XB=0.0,25.0,-20.0,20.0,0.0,45.0, MPI_PROCESS=11/

&SPEC ID='WATER VAPOR',DENSITY_LIQUID=1000,VAPORIZATION_TEMPERATURE=100/

&PART ID='Water_droplets',
HEAT_TRANSFER_COEFFICIENT_GAS=0,
MASS_TRANSFER_COEFFICIENT=0,
DIAMETER=82,
MONODISPERSE=T,
PRIMARY_BREAKUP_LENGTH=10, PRIMARY_BREAKUP_DRAG_REDUCTION_FACTOR=0.01,
SPEC_ID='WATER VAPOR',
QUANTITIES='PARTICLE DIAMETER','PARTICLE VELOCITY'/

&SURF ID='Burner', COLOR='RED', PART_ID='Water_droplets', SPEC_ID='WATER VAPOR', PARTICLE_MASS_FLUX=10600, VEL=0, VEL_PART=-285, RAMP_PART='Burner_RAMP_Q',DT_INSERT=0.005, NPPC=25/

&RAMP ID='Burner_RAMP_Q', T=0.0, F=0.0/
&RAMP ID='Burner_RAMP_Q', T=0.001, F=0.1/
&RAMP ID='Burner_RAMP_Q', T=0.002, F=0.2/
&RAMP ID='Burner_RAMP_Q', T=0.003, F=0.3/
&RAMP ID='Burner_RAMP_Q', T=0.004, F=0.4/
&RAMP ID='Burner_RAMP_Q', T=0.005, F=0.5/
&RAMP ID='Burner_RAMP_Q', T=0.006, F=0.6/
&RAMP ID='Burner_RAMP_Q', T=0.007, F=0.7/
&RAMP ID='Burner_RAMP_Q', T=0.008, F=0.8/
&RAMP ID='Burner_RAMP_Q', T=0.009, F=0.9/
&RAMP ID='Burner_RAMP_Q', T=0.01, F=1.0/
&RAMP ID='Burner_RAMP_Q', T=0.08, F=1.0/
&RAMP ID='Burner_RAMP_Q', T=0.081, F=0.9/
&RAMP ID='Burner_RAMP_Q', T=0.082, F=0.8/
&RAMP ID='Burner_RAMP_Q', T=0.083, F=0.7/
&RAMP ID='Burner_RAMP_Q', T=0.084, F=0.6/
&RAMP ID='Burner_RAMP_Q', T=0.085, F=0.5/
&RAMP ID='Burner_RAMP_Q', T=0.086, F=0.4/
&RAMP ID='Burner_RAMP_Q', T=0.087, F=0.3/
&RAMP ID='Burner_RAMP_Q', T=0.088, F=0.2/
&RAMP ID='Burner_RAMP_Q', T=0.089, F=0.1/
&RAMP ID='Burner_RAMP_Q', T=0.09, F=0.0/
&RAMP ID='Burner_RAMP_Q', T=10, F=0.0/

&OBST XB=3.0,8.0,-12,12,0.0,5.0, SURF_ID='INERT'/
&OBST XB=8.0,12.0,-4,4,0.0,12.0, SURF_ID='INERT'/
&OBST XB=15.0,50.0,-60,60,0.0,2.0, SURF_ID='INERT'/

&GEOM XB=23.0,27.0,-2.0,2.0,0.0,5, SURF_IDS='Burner','INERT','INERT' /

&VENT ID='Mesh Vent: Mesh [XMAX]', SURF_ID='OPEN', XB=50.0,50.0,-60.0,60.0,0.0,90/
&VENT ID='Mesh Vent: Mesh [XMIN]', SURF_ID='OPEN', XB=0.0,0.0,-60.0,60.0,0.0,90/
&VENT ID='Mesh Vent: Mesh [YMAX]', SURF_ID='OPEN', XB=0.0,50.0,60.0,60.0,0.0,90.0/
&VENT ID='Mesh Vent: Mesh [YMIN]', SURF_ID='OPEN', XB=0.0,50.0,-60.0,-60.0,0.0,90.0/
&VENT ID='Mesh Vent: Mesh [ZMAX]', SURF_ID='OPEN', XB=0.0,50.0,-60.0,60.0,90.0,90.0/
&VENT ID='Mesh Vent: Mesh [ZMIN]',
SURF_ID='INERT', XB=0.0,50.0,-60.0,60.0,0.0,0.0/

&DEVC ID='TotalParticleMass', QUANTITY='MPUV', PART_ID='Water_droplets',
XB=0.0,45.0,-60.0,60.0,0.0,90.0, SPATIAL_STATISTIC='VOLUME INTEGRAL' /
&DEVC ID='InstantaneousSurfaceMass', QUANTITY='MPUA', PART_ID='Water_droplets', XB=0.0,45.0,-60.0,60.0,0.0,0.0, SPATIAL_STATISTIC='SURFACE INTEGRAL' /
&DEVC ID='Divergence', QUANTITY='DIVERGENCE', XB=0.0,45.0,-60.0,60.0,0.0,90.0, SPATIAL_STATISTIC='MAX' /
&SLCF PBY=0.0, QUANTITY='MASS FRACTION', SPEC_ID='WATER VAPOR' /
&SLCF PBY=0.0, QUANTITY='VELOCITY' /
&TAIL

[marcosvanella]

@mohanvadaga
I would like to know whether fds supports SURF IDs with VEL_PART on GEOM IDs
You hit the nail in the head. VEL_PART was not included as insertion velocity option in geometry faces. See PR #15714.
If you are running from a bundle, you'll need to wait until the next nightly with the change is here. Otherwise update your repo and recompile fds.

[mohanvadaga]

Thank you,Marco for the quick response. This is not related to the question I posted, I would like to understand, are there any other options for droplet releases. Because I was going through this paper, "https://doi.org/10.1016/j.nucengdes.2017.04.012", where they have released droplets through spherical surface refer to fig.1&2. How can we do this, as it was done in 2017, havent they used GEOM ID. I understand that multiple point sources using INIT can also be used to represent radial position and radial velocity vectors with some pre processing. But with that option we will not be able to represent particle parcels(i mean each parcel will represent one droplet) or distribution or accurately give mass flux.

[mohanvadaga]

Dear Marco, I have tried running the latest NIGHTY file still facing the same problem. The code and particle velocity are shown below:

With GEOM:
&HEAD CHID="Sandia_Phantom_Exp"/
&TIME T_END=5, DT=0.001/
&DUMP DT_RESTART=5, NFRAMES=1000,DT_DEVC=0.001,VELOCITY_ERROR_FILE=.TRUE.,MAXIMUM_PARTICLES=10000000/
&MISC SIMULATION_MODE='LES', TURBULENCE_MODEL='DEARDORFF',POROUS_FLOOR=F,CFL_MAX=0.1, PARTICLE_CFL_MAX=0.1,PARTICLE_CFL=.TRUE./
&PRES SUSPEND_PRESSURE_ITERATIONS=.TRUE.,MAX_PRESSURE_ITERATIONS=50,PRESSURE_TOLERANCE=0.00000001,VELOCITY_TOLERANCE=0.00000001,SOLVER='UGLMAT'/

&MESH ID='Mesh-1', IJK=20,40,50, XB=0.0,20.0,-20.0,20.0,0.0,50.0/

&SPEC ID='WATER VAPOR',DENSITY_LIQUID=0.1,VAPORIZATION_TEMPERATURE=1500/

&PART ID='Water_droplets',
HEAT_TRANSFER_COEFFICIENT_GAS=0,
MASS_TRANSFER_COEFFICIENT=0,
DIAMETER=82,
MONODISPERSE=T,
SPEC_ID='WATER VAPOR',
N_STRATA=3,
CHECK_DISTRIBUTION=T,
PRIMARY_BREAKUP_LENGTH=10,
PRIMARY_BREAKUP_DRAG_REDUCTION_FACTOR=0.0,
QUANTITIES='PARTICLE DIAMETER','PARTICLE VELOCITY'/

&SURF ID='Burner', COLOR='RED', PART_ID='Water_droplets', SPEC_ID='WATER VAPOR', PARTICLE_MASS_FLUX=100, VEL=0, VEL_PART=-30,RAMP_PART='Burner_RAMP_Q',DT_INSERT=0.005, NPPC=3/

&RAMP ID='Burner_RAMP_Q', T=0.0, F=0.0/
&RAMP ID='Burner_RAMP_Q', T=0.01, F=1.0/
&RAMP ID='BurnerRAMP_Q', T=0.08, F=1.0/
&RAMP ID='Burner_RAMP_Q', T=0.09, F=0.0/
&RAMP ID='Burner_RAMP_Q', T=10, F=0.0/

&GEOM XB=9.0,11.0,-1.0,1.0,0.0,2.1, SURF_IDS='Burner','INERT','INERT'/

&VENT ID='Mesh Vent: Mesh [XMAX]', SURF_ID='OPEN', XB=20.0,20.0,-20.0,20.0,0.0,50/
&VENT ID='Mesh Vent: Mesh [XMIN]', SURF_ID='OPEN', XB=0.0,0.0,-20.0,20.0,0.0,50/
&VENT ID='Mesh Vent: Mesh [YMAX]', SURF_ID='OPEN', XB=0.0,20.0,20.0,20.0,0.0,50.0/
&VENT ID='Mesh Vent: Mesh [YMIN]', SURF_ID='OPEN', XB=0.0,20.0,-20.0,-20.0,0.0,50.0/
&VENT ID='Mesh Vent: Mesh [ZMAX]', SURF_ID='OPEN', XB=0.0,20.0,-20.0,20.0,50.0,50.0/
&VENT ID='Mesh Vent: Mesh [ZMIN]', SURF_ID='INERT', XB=0.0,20.0,-20.0,20.0,0.0,0.0/

&DEVC ID='TotalParticleMass', QUANTITY='MPUV', PART_ID='Water_droplets',
XB=0.0,20.0,-20.0,20.0,0.0,50.0, SPATIAL_STATISTIC='VOLUME INTEGRAL' /
&DEVC ID='InstantaneousSurfaceMass', QUANTITY='MPUA', PART_ID='Water_droplets', XB=0.0,20.0,-20.0,20.0,0.0,0.0, SPATIAL_STATISTIC='SURFACE INTEGRAL' /
&DEVC ID='Divergence', QUANTITY='DIVERGENCE', XB=0.0,20.0,-20.0,20.0,0.0,50.0, SPATIAL_STATISTIC='MAX' /
&SLCF PBY=0.0, QUANTITY='MASS FRACTION', SPEC_ID='WATER VAPOR' /
&SLCF PBY=0.0, QUANTITY='VELOCITY' /
&TAIL/

With GEOM you can see that the velocity is dropping suddenly:

While with OBST the velocity is seen to be close to 30m/s. The small drop may be attributed to gravity:

[marcosvanella]

We had a couple of fails on our CI system firebot over the past days. The latest revision that passed the verification suite is from before my change on Friday. Keep an eye here for a build sucess, to get the latest source on the nightly bundle (I'll also let you know). Otherwise you will have to compile fds from latest repository.

On the method to eject droplets at high speed you cite, maybe @shostikk can provide some information.