I can not get the desired porosity with my PSD

Asked by Saeed on 2019-06-02

Hello All,
I am trying to generate sphere packing according to a given PSD.

I want to get a porosity=0.375 or e=(triax.porosity)/(1-triax.porosity)=0.6

I use num_spheres=1000 and my answer for porosity is 0.4969
I use num_spheres=5000 and my answer for porosity is 0.435
I use num_spheres=15000 and my answer for porosity is 0.411
I use num_spheres=20000 and my answer for porosity is 0.4139
I use num_spheres=30000 and my answer for porosity is 0.4059
I use num_spheres=40000 and my answer for porosity is 0.39
I use num_spheres=50000 and my answer for porosity is 0.39
I use num_spheres=60000 and my answer for porosity is 0.39

How to use this psd to get the desired porosity=0.375 or e=(triax.porosity)/(1-triax.porosity)=0.6 ?

=============================================================
import matplotlib; matplotlib.rc('axes',grid=True)
from yade import pack
import pylab
from numpy import *
import matplotlib.pyplot as plt
import math

from yade.params import table

seed=table.seed
num_spheres=table.num_spheres# number of spheres
compFricDegree = table.compFricDegree # initial contact friction during the confining phase (will be decreased during the REFD compaction process)
confiningS=-2820

## creat a packing with a specific particle side distribution (PSD)
psdSizes,psdCumm=[0.0011,0.0012,0.0030,0.0059,0.0081,0.0111,0.0133,0.0222,0.0285,0.0371,0.0746,0.1491,0.2982,0.4224],[0.,0.22,0.29,0.36,.45,0.55,0.59,0.69,0.76,0.87,0.94,0.95,0.97,1.0]
sp=pack.SpherePack()
mn,mx=Vector3(0,0,0),Vector3(2.5,2.5,2.5)
sp.makeCloud(minCorner=mn,maxCorner=mx,num=num_spheres,psdSizes=psdSizes,psdCumm=psdCumm,distributeMass=False,seed=seed)
sp.psd(bins=5000,mass=False)
sp.toSimulation

## create material #0, which will be used as default
O.materials.append(FrictMat(young=15e6,poisson=.4,frictionAngle=0,density=0,label='frictionless'))

## create walls around the packing
walls=aabbWalls((mn,mx),thickness=0,material='frictionless')
wallIds=O.bodies.append(walls)

triax=TriaxialStressController(
internalCompaction=False,
goal1=confiningS,
goal2=confiningS,
goal3=confiningS,
max_vel=10,
label="triax"
)

newton=NewtonIntegrator(damping=0.4)

O.engines=[
ForceResetter(),
InsertionSortCollider([Bo1_Sphere_Aabb(),Bo1_Box_Aabb()]),
InteractionLoop(
[Ig2_Sphere_Sphere_ScGeom(),Ig2_Box_Sphere_ScGeom()],
[Ip2_FrictMat_FrictMat_FrictPhys()],
[Law2_ScGeom_FrictPhys_CundallStrack()]
),
GlobalStiffnessTimeStepper(active=1,timeStepUpdateInterval=100,timestepSafetyCoefficient=0.8),
triax,
newton
]

O.dt=.1*PWaveTimeStep()
O.dynDt=False

while 1:
O.run(1000, True)
unb=unbalancedForce()
print 'unb: ',unb,' t.p: ',triax.porosity,' n: ',porosity(),' e: ',(triax.porosity)/(1-triax.porosity),' Sigma: ',-triax.stress(3)
if unb<0.001 and abs(confiningS-triax.meanStress)/(-confiningS)<0.001:
break

print 'porosity: ',porosity()
print 'triax.porosity',triax.porosity
print 'e: ',(triax.porosity)/(1-triax.porosity)
==========================================================

Question information

Language:
English Edit question
Status:
For:
Assignee:
No assignee Edit question
Last query:
2019-06-06
2019-06-06
 Robert Caulk (rcaulk) said on 2019-06-03: #1

Have you tried changing the PSD to increase the fraction of smaller particles? Have you tried not passing numSpheres to makeCloud? 

 Saeed (saeed0668191) said on 2019-06-06: #2

hi Robert
I increase the number of compFricDegree, but porosity (n) and Void Ratio: e=n/(1-n) increased and did not help.
Without using the num_spheres in sp.makeCloud , it simulates the small interval of the psd, and the entire interval of the seeds does not coat, and a very large amount of grain is modeled, which is very long runtime.
And the psd does not fully model the numbers given in psdSizes and psdCumm.
Please provide a solution
thanks

 Robert Caulk (rcaulk) said on 2019-06-06: #3

Have you tried changing the PSD to increase the fraction of smaller particles?

Have you tried following the triax-example porosity method ?

 Bruno Chareyre (bruno-chareyre) said on 2019-06-06: #4

distributeMass=True would probably help