Dear all,
I am a beginner in LIGGGHTS
I am trying to simulate the sediment transport around the pile based
on the reference paper "Shim, J., Duan, J., & Jo, H. (2016).
Simulating Sediment Transport around a Bridge Pier Using Open FOAM
Software. In World Environmental and Water Resources Congress 2016
(pp. 362-369)".
This paper validates by comparing between the simulated the sediment transport and experiment around the pier using CFDEM coupling.
Reynolds number is about 1.8e4. it is laminar flow.
I modified the tutorial case "ErgunTestMPI".
First of all, the particles fly with a light feeling as shown attached file.
The part of the run script is as follow,
atom_style granular
atom_modify map array
communicate single vel yes
soft_particles yes
boundary ff pp ff
newton off
units si
region reg block -0.0875 0.0875 -0.0375 0.0375 -0.001 0.06 units box
create_box 1 reg
neighbor 0.001 bin
neigh_modify delay 0
# Material properties required for granular pair styles
fix m1 all property/global youngsModulus peratomtype 5.e5 #5.e6 #5.e6
fix m2 all property/global poissonsRatio peratomtype 0.45
fix m3 all property/global coefficientRestitution peratomtypepair 1 0.051 #0.3 it must be higher than 0.05
fix m4 all property/global coefficientFriction peratomtypepair 1 0.6 #0.5
# pair style
pair_style gran model hertz tangential history
pair_coeff * *
# timestep, gravity
timestep 0.00001
fix gravi all gravity 9.81 vector 0.0 0.0 -1.0
fix xwalls1 all wall/gran model hertz tangential history primitive type 1 xplane -0.0875
fix xwalls2 all wall/gran model hertz tangential history primitive type 1 xplane 0.0875
fix ywalls1 all wall/gran model hertz tangential history primitive type 1 yplane -0.0375
fix ywalls2 all wall/gran model hertz tangential history primitive type 1 yplane 0.0375
fix zwalls1 all wall/gran model hertz tangential history primitive type 1 zplane -0.001
fix zwalls2 all wall/gran model hertz tangential history primitive type 1 zplane 0.06
fix Cylinder all wall/gran model hertz tangential history primitive type 1 zcylinder 0.011 0. 0.
#group nve_group region reg
region bc cylinder z 0.0 0.0 0.011 -0.001 0.06 units box side out
region domain intersect 2 reg bc
region flat block -0.0875 0.0875 -0.0375 0.0375 0.03 0.06 units box
# particle distributions and insertion
#region bc cylinder z 0.0 0.0 0.012 0. 0.055 units box
fix pts1 all particletemplate/sphere 15485863 atom_type 1 density constant 2650 radius constant 0.0012
fix pdd1 all particledistribution/discrete 15485867 1 pts1 1.0
fix ins all insert/pack seed 32452843 distributiontemplate pdd1 vel constant 0. 0. -1. insert_every once overlapcheck yes all_in yes
particles_in_region 50000 region domain
# apply nve integration to all particles that are inserted as single particles
fix integr all nve/sphere
The questions are as follow,
1. Which particle model should be used to implement the seabed? I have searched a lot of references but I am not sure which model is suitable
because I am not familiar with DEM.
2. In the reference paper, the restitution coefficient is 0.01. inside the code, however, it must be above 0.05. is there a way to set this value to 0.01?
3. I want to build the particle up to seabed level, like a flatbed. So, first, enough particles are deposited higher than seabed level and
then deleted at the seabed level using delete_atoms command. however, it is not the perfect flatbed.
4. I set the rho value to 1000 in the CFD/0 directory. However, I know that it is not necessary to set the density in incompressible flows.
So where is this rho value used in cfdem coupling?
5. How do I set the inlet velocity? First, I divided the inlet patch into two patches(flow patch and seabed patch) based on the seabed level.
and a uniform value or the fully developed laminar boundary velocity profile using groovyBC was set on the flow inlet patch. is it right?
I have searched a lot for this issue, but it is not easy to apply to these.
Any answers or advice are welcome.
Any insights/comments would be appreciated. Thanks.
| Attachment | Size |
|---|---|
 flying_particle.png | 296.98 KB |
paul | Sat, 04/07/2018 - 08:43
1. Hertz is a good first
1. Hertz is a good first start. I do not know how important rolling friction is for your application.
2. There will be virtually no difference with a restitution coefficient that low. Particles will end up with basically no relative velocity after colliding.
3. Doing some compression using a servo wall may be useful for that problem.
4. rho is needed for the drag force calculation.
5. Don't know, post 0/U so we can have a look at it. Your description sounds reasonable though.
Your simulation domain looks very small to me - do the results change if you choose a bigger domain?
Since you have a liquid phase: Have you enabled the virtual mass force?
Minghao | Mon, 04/09/2018 - 20:24
Hi ssong,
Hi ssong,
I have the same interests on this topic. I will follow the relative updates.
ssong | Wed, 04/11/2018 - 14:41
Many thanks, Paul.
Many thanks, Paul,
I tested some cases based on your advice.
The results are better. there are no flying particles.
But the depth of scour does not appear sufficiently.
I changed several things.
First, material properties have been changed based on Ergun tutorials.
Second, the size of the domain was extended to be the same as the size of the reference paper.
Next, the value of the neighbor set was changed to the particle diameter.
From now on, I think I need to validate my result using the experiment
data by other authors
I have some questions.
1. how to plot or measure the scour depth in paraview in order to compare with an experiment data. In case of the wave pattern on the ship, alpha_water value is used.
2. how to decide the proper material properties when I describe the seabed. Is there any guide paper or reference?
If you give me any advice and guide, it will be very helpful.
Thank you in advance.
#hi, Minghao. we can share some helpful information. thank you!