FuncG in ConcretePM.cpp

Hello Hicham,
when kappa==epsCrackOnset (eps_0), the damage should be 0, so there is a
misprint in the thesis and the code is correct.
cheers
Jan

2013/12/2 Hicham BENNIOU <email address hidden>

> Question #240167 on Yade changed:
> https://answers.launchpad.net/yade/+question/240167
>
> Description changed to:
> Hi everyone,
>
> This question is a bit particular because only Vaclav Smilauer or a person
> familiar with the code can answer.
> I was comparing the model described in Smilauer's Thesis [1] to the code
> ConcretePM.cpp [2], in line 158 I read that funcG returns:
>
> 1.-(epsCrackOnset/kappaD)*exp(-(kappaD-epsCrackOnset)/epsFracture)
>
> Or, in the Thesis, page 38 I read that funcG is:
>
> 1.-(epsFracture/kappa)*exp(-(kappa-epsCrackOnset)/epsFracture)
>
> knowing that kappaD = max (epsilonN) and epsFracture=30*epsCrackOnset ,
> I'm a bit confused here.
>
> Any explanation would be welcome
>
> Thanks !
>
> [1] http://beta.arcig.cz/~eudoxos/smilauer2010-phd-thesis.pdf
> [2] https://github.com/hbenniou/trunk/blob/master/pkg/dem/ConcretePM.cpp
>
> --
> You received this question notification because you are a member of
> yade-users, which is an answer contact for Yade.
>
> _______________________________________________
> Mailing list: https://launchpad.net/~yade-users
> Post to : <email address hidden>
> Unsubscribe : https://launchpad.net/~yade-users
> More help : https://help.launchpad.net/ListHelp
>

Thanks Jan Stránský, that solved my question.

Hello,

Bringing this subject again cause I don't want to flood the forum with topics.

Is it possible to use Law2_ScGeom_CpmPhys_cpm() with TriaxialStressController() ? I'm trying to run a script I wrote using this constitutive law with this engine but i'm getting a lot of errors.

thanks !

Hi Hicham,
please attach the errors you are getting and also a minimal working script
if it is not too long.
thanks
Jan

2013/12/4 Hicham BENNIOU <email address hidden>

> Question #240167 on Yade changed:
> https://answers.launchpad.net/yade/+question/240167
>
> Status: Solved => Open
>
> Hicham BENNIOU is still having a problem:
> Hello,
>
> Bringing this subject again cause I don't want to flood the forum with
> topics.
>
> Is it possible to use Law2_ScGeom_CpmPhys_cpm() with
> TriaxialStressController() ? I'm trying to run a script I wrote using
> this constitutive law with this engine but i'm getting a lot of errors.
>
> thanks !
>
> --
> You received this question notification because you are a member of
> yade-users, which is an answer contact for Yade.
>
> _______________________________________________
> Mailing list: https://launchpad.net/~yade-users
> Post to : <email address hidden>
> Unsubscribe : https://launchpad.net/~yade-users
> More help : https://help.launchpad.net/ListHelp
>

Hicham, please open a new question if it is unrelated.

B

Here is the script :

from yade import pack, qt, plot

# Vars definition
specimenLength = 15.0
spheresInCell = 2000
sphereRadius= 3.5e-3
intRadius = 1.5 #Interaction Radius

sigmaT=3.5e6,
epsCrackOnset=1e-4,
relDuctility=30,
isoPrestress=-50000000, #isotropic confinment

strainRateTension=0.5,
strainRateCompression=.5,

youngConcrete = 24e9
youngWalls = 0.5

poissonConcrete = 0.2
poissonWalls = 0.5

frictionAngleConcrete = atan(0.8)
frictionAngleWalls = 0

densityConcrete = 4800
densityWalls = 0

damping=0.4

#Create walls around the packing of spheres
min = Vector3(-.5*specimenLength,-.5*specimenLength,-.5*specimenLength)
max = Vector3(.5*specimenLength,.5*specimenLength,.5*specimenLength)

Walls =aabbWalls([min,max], thickness = 0)
wallIds = O.bodies.append(Walls)

#Definig Concrete material and packing spheres
concreteId=O.materials.append(CpmMat(young=youngConcrete,frictionAngle=frictionAngleConcrete,poisson=poissonConcrete,density=densityConcrete,sigmaT=sigmaT,relDuctility=relDuctility,epsCrackOnset=epsCrackOnset,isoPrestress=isoPrestress))
sp=pack.randomDensePack(pack.inAlignedBox(min,max),spheresInCell=spheresInCell,radius=sphereRadius,returnSpherePack=True)
sp.toSimulation(material=concreteId)

# Defining Engines
TriaxEngine = TriaxialStressController(
      maxMultiplier=1.+2e4/youngConcrete,
      finalMaxMultiplier=1.+2e3/youngConcrete,
      thickness = 0,
      stressMask = 7, # 7=111, 1->x, 1->y and 1->z, which means stress is controlled on xyz
      internalCompaction=True,
)

O.engines=[
 ForceResetter(),
 InsertionSortCollider([Bo1_Sphere_Aabb(),Bo1_Box_Aabb()],verletDist=.05*sphereRadius),
 InteractionLoop(
  [Ig2_Sphere_Sphere_ScGeom(interactionDetectionFactor=intRadius),Ig2_Box_Sphere_ScGeom()],
  [Ip2_CpmMat_CpmMat_CpmPhys()],
  [Law2_ScGeom_CpmPhys_Cpm()],
 ),

 TriaxEngine,
 GlobalStiffnessTimeStepper(active=1,timeStepUpdateInterval=100,timestepSafetyCoefficient=0.8),
 PyRunner(iterPeriod=10,initRun=True,command='history()'),
 TriaxialStateRecorder(iterPeriod=100,file='~/Bureau/WallStresses'),

 NewtonIntegrator(damping=damping,gravity=(0.,0.,-10.)),
]

# record and plot data

def history():
  plot.addData(e11=TriaxEngine.strain[0], e22=TriaxEngine.strain[1], e33=TriaxEngine.strain[2],
  ev=-TriaxEngine.strain[0]-TriaxEngine.strain[1]-TriaxEngine.strain[2],
  s11=TriaxEngine.stress(TriaxEngine.wall_right_id)[0],
  s22=TriaxEngine.stress(TriaxEngine.wall_top_id)[1],
  s33=TriaxEngine.stress(TriaxEngine.wall_front_id)[2],
  i=O.iter
 )

O.run(10000)

plot.plots={'e22':('s11','s22','s33','ev')}
plot.plot()

Hi Hicham,

the first problem is in material parameters definition. If you use comma at
the end of line, the variable is tuple instead of float. If you want to
pass it to CpmMat constructor, it expects number but gets tuple and raise
the error.

you can try in IPython
a = 1.4,
a.__class__.__name__ # returns 'tuple'
a = 1.4 # without comma
a.__class__.__name__ # returns 'float'

so use
sigmaT = 3.5e6
instead of
sigmaT = 3.5e6,

and the same also for all other material parameters

cheers
Jan

2013/12/4 Hicham BENNIOU <email address hidden>

> Question #240167 on Yade changed:
> https://answers.launchpad.net/yade/+question/240167
>
> Status: Answered => Open
>
> Hicham BENNIOU is still having a problem:
> Here is the script :
>
> from yade import pack, qt, plot
>
> # Vars definition
> specimenLength = 15.0
> spheresInCell = 2000
> sphereRadius= 3.5e-3
> intRadius = 1.5 #Interaction Radius
>
> sigmaT=3.5e6,
> epsCrackOnset=1e-4,
> relDuctility=30,
> isoPrestress=-50000000, #isotropic confinment
>
> strainRateTension=0.5,
> strainRateCompression=.5,
>
> youngConcrete = 24e9
> youngWalls = 0.5
>
> poissonConcrete = 0.2
> poissonWalls = 0.5
>
> frictionAngleConcrete = atan(0.8)
> frictionAngleWalls = 0
>
> densityConcrete = 4800
> densityWalls = 0
>
> damping=0.4
>
> #Create walls around the packing of spheres
> min = Vector3(-.5*specimenLength,-.5*specimenLength,-.5*specimenLength)
> max = Vector3(.5*specimenLength,.5*specimenLength,.5*specimenLength)
>
> Walls =aabbWalls([min,max], thickness = 0)
> wallIds = O.bodies.append(Walls)
>
> #Definig Concrete material and packing spheres
>
> concreteId=O.materials.append(CpmMat(young=youngConcrete,frictionAngle=frictionAngleConcrete,poisson=poissonConcrete,density=densityConcrete,sigmaT=sigmaT,relDuctility=relDuctility,epsCrackOnset=epsCrackOnset,isoPrestress=isoPrestress))
>
> sp=pack.randomDensePack(pack.inAlignedBox(min,max),spheresInCell=spheresInCell,radius=sphereRadius,returnSpherePack=True)
> sp.toSimulation(material=concreteId)
>
> # Defining Engines
> TriaxEngine = TriaxialStressController(
> maxMultiplier=1.+2e4/youngConcrete,
> finalMaxMultiplier=1.+2e3/youngConcrete,
> thickness = 0,
> stressMask = 7, # 7=111, 1->x, 1->y and 1->z, which means stress is
> controlled on xyz
> internalCompaction=True,
> )
>
> O.engines=[
> ForceResetter(),
>
> InsertionSortCollider([Bo1_Sphere_Aabb(),Bo1_Box_Aabb()],verletDist=.05*sphereRadius),
> InteractionLoop(
>
> [Ig2_Sphere_Sphere_ScGeom(interactionDetectionFactor=intRadius),Ig2_Box_Sphere_ScGeom()],
> [Ip2_CpmMat_CpmMat_CpmPhys()],
> [Law2_ScGeom_CpmPhys_Cpm()],
> ),
>
> TriaxEngine,
>
> GlobalStiffnessTimeStepper(active=1,timeStepUpdateInterval=100,timestepSafetyCoefficient=0.8),
> PyRunner(iterPeriod=10,initRun=True,command='history()'),
> TriaxialStateRecorder(iterPeriod=100,file='~/Bureau/WallStresses'),
>
> NewtonIntegrator(damping=damping,gravity=(0.,0.,-10.)),
> ]
>
> # record and plot data
>
> def history():
> plot.addData(e11=TriaxEngine.strain[0], e22=TriaxEngine.strain[1],
> e33=TriaxEngine.strain[2],
>
> ev=-TriaxEngine.strain[0]-TriaxEngine.strain[1]-TriaxEngine.strain[2],
> s11=TriaxEngine.stress(TriaxEngine.wall_right_id)[0],
> s22=TriaxEngine.stress(TriaxEngine.wall_top_id)[1],
> s33=TriaxEngine.stress(TriaxEngine.wall_front_id)[2],
> i=O.iter
> )
>
> O.run(10000)
>
> plot.plots={'e22':('s11','s22','s33','ev')}
> plot.plot()
>
> --
> You received this question notification because you are a member of
> yade-users, which is an answer contact for Yade.
>
> _______________________________________________
> Mailing list: https://launchpad.net/~yade-users
> Post to : <email address hidden>
> Unsubscribe : https://launchpad.net/~yade-users
> More help : https://help.launchpad.net/ListHelp
>

Thanks Jan,

It still generates this error :

sp=pack.randomDensePack(pack.inAlignedBox((-.5*specimenLength,-.5*specimenLength,-.5*specimenLength),(.5*specimenLength,.5*specimenLength,.5*specimenLength)),spheresInCell=spheresInCell,radius=sphereRadius,returnSpherePack=True)
  File "~/Yade/build/lib/x86_64-linux-gnu/yade-2013-11-20.git-11c208a/py/yade/pack.py", line 474, in randomDensePack
    if wantPeri: sp.cellFill(Vector3(fullDim[0],fullDim[1],fullDim[2]))
MemoryError

This is weird especially because I use the same line that generates this error in another script and it works fine :
sp=pack.randomDensePack(pack.inAlignedBox(min,max),spheresInCell=spheresInCell,radius=sphereRadius,returnSpherePack=True)

Any Ideas ?

Hicham

Hi Hicham,
check if not only the line, but also specimenLength and sphereRadius values
are the same (or similar). According to MemoryError I would say that
the specimenLength/sphereRadius ratio is much higher in the error prone
simulation than that you used before, resulting in too many particles
exceeding memory capacity.
cheers
Jan

2013/12/4 Hicham BENNIOU <email address hidden>

> Question #240167 on Yade changed:
> https://answers.launchpad.net/yade/+question/240167
>
> Status: Answered => Open
>
> Hicham BENNIOU is still having a problem:
> Thanks Jan,
>
> It still generates this error :
>
>
> sp=pack.randomDensePack(pack.inAlignedBox((-.5*specimenLength,-.5*specimenLength,-.5*specimenLength),(.5*specimenLength,.5*specimenLength,.5*specimenLength)),spheresInCell=spheresInCell,radius=sphereRadius,returnSpherePack=True)
> File
> "~/Yade/build/lib/x86_64-linux-gnu/yade-2013-11-20.git-11c208a/py/yade/pack.py",
> line 474, in randomDensePack
> if wantPeri: sp.cellFill(Vector3(fullDim[0],fullDim[1],fullDim[2]))
> MemoryError
>
> This is weird especially because I use the same line that generates this
> error in another script and it works fine :
>
> sp=pack.randomDensePack(pack.inAlignedBox(min,max),spheresInCell=spheresInCell,radius=sphereRadius,returnSpherePack=True)
>
> Any Ideas ?
>
> Hicham
>
> --
> You received this question notification because you are a member of
> yade-users, which is an answer contact for Yade.
>
> _______________________________________________
> Mailing list: https://launchpad.net/~yade-users
> Post to : <email address hidden>
> Unsubscribe : https://launchpad.net/~yade-users
> More help : https://help.launchpad.net/ListHelp
>

Hey, seeing this thread late, I want to apologize for the typo in the thesis. Good job spotting that! Cheers, v.