I need to simulate uniaxial compaction

In the name of God
Hi Nima

If you are familiar with triaxialStressController engine, then you easily can deactivate all walls and then try generating and compacting each layer of grains.
compaction can be applied with a plate.

Regards
Mohsen

 Hi Mohsen,
Thanks so much. Have you done this before? if yes, would you please share this part of the code with me if possible? I basically don't know why we need to  use triaxialStressController to perform the compaction. We are not in the experiment phase yet. This is a kind of pre-experiment like we actually do in the lab, compacting different layers and measuring void ratio after each compaction. I know that I need to write a loop to to reach to the target void ratio of each compaction but have some difficulties with understanding triaxialStressController for this case. Do you think that I need to set:triax.goal1=triax.goal2=triax.goal3=0

first

and then apply the vertical pressure on top of a rigid plate to compact the layer. If required I can send you my script to take a look and modify.

Thanks sincerely for your kind attention.
Nima
    On Thursday, October 5, 2017, 6:28:02 AM PDT, mohsen <email address hidden> wrote:

 Your question #658905 on Yade changed:
https://answers.launchpad.net/yade/+question/658905

    Status: Open => Answered

mohsen proposed the following answer:
In the name of God
Hi Nima

If you are familiar with triaxialStressController engine, then you easily can deactivate all walls and then try generating and compacting each layer of grains.
compaction can be applied with a plate.

Regards
Mohsen

--
If this answers your question, please go to the following page to let us
know that it is solved:
https://answers.launchpad.net/yade/+question/658905/+confirm?answer_id=0

If you still need help, you can reply to this email or go to the
following page to enter your feedback:
https://answers.launchpad.net/yade/+question/658905

You received this question notification because you asked the question.

i have not done exactly the same thing, but you can generate your sample in a box (look at Oedometer example in tutorial) layer by layer and then save and load the grains (or grain positions and diameters) to do a triaxial shearing; so basically triaxial engine is not the only way!

Note that by deactivating the triaxial walls, they would not have any movement:
https://yade-dem.org/doc/yade.wrapper.html?highlight=triaxialstresscontroller#yade.wrapper.TriaxialStressController.wall_back_activated

Regards
Mohsen

Hi Mohsen,
Thanks again for your hint. Do you mean I need to have two distinct scripts if I want to follow the instructions like in odeometer test? I don't need an exact script for what I am doing, even if you have an approximate solution (script) regarding deactivating walls that would be enough for me. Another critical question is about how recognizing the surface of previous layers. I mean, I am compacting layers based on the geometry before compaction. How we can enforce the code to understand the new position of layers (top surface of layers will be lower after compaction but I am generating the next layer exactly from the original height of top surface of previous layer  (before compaction)). Is there any solution?
Sincerely,
Nima On Thursday, October 5, 2017, 8:17:34 PM PDT, mohsen <email address hidden> wrote:

 Your question #658905 on Yade changed:
https://answers.launchpad.net/yade/+question/658905

    Status: Open => Answered

mohsen proposed the following answer:
i have not done exactly the same thing, but you can generate your sample
in a box (look at Oedometer example in tutorial)  layer by layer and
then save and load the grains (or grain positions and diameters) to do a
triaxial shearing; so basically triaxial engine is not the only way!

Note that by deactivating the triaxial walls, they would not have any movement:
https://yade-dem.org/doc/yade.wrapper.html?highlight=triaxialstresscontroller#yade.wrapper.TriaxialStressController.wall_back_activated

Regards
Mohsen

--
If this answers your question, please go to the following page to let us
know that it is solved:
https://answers.launchpad.net/yade/+question/658905/+confirm?answer_id=2

If you still need help, you can reply to this email or go to the
following page to enter your feedback:
https://answers.launchpad.net/yade/+question/658905

You received this question notification because you asked the question.

>>Thanks again for your hint. Do you mean I need to have two distinct scripts if I want to follow the instructions like in odeometer test?

Nima If you follow this, you need to scripts. the first one is for sample generation and compaction; and the second one for shearing. i do not have any script regarding this approach.

>> I don't need an exact script for what I am doing, even if you have an approximate solution (script) regarding deactivating walls that would be enough for me.

You can easily deactivates walls:

wall_front_activated=False

there are 6 walls which in your case all of them should be deactivate. you may define a plate for compaction. After compaction of each layer you may move or delete the plate and then you can generate again the next layer and compact and so on

>Another critical question is about how recognizing the surface of previous layers. I mean, I am compacting layers based on the geometry before compaction. How we can enforce the code to understand the new position of layers (top surface of layers will be lower after compaction but I am generating the next layer exactly from the original height of top surface of previous layer (before compaction)). Is there any solution?

You easily can distinguish the new position of the recent layer that has been just compacted:

MaxZ=max([b.state.pos[2]+b.shape.radius for b in O.bodies if isinstance(b.shape,Sphere)])
MaxZ is the coordination of the grain with maximum height (Note that Z is defined as the axis which grain are compacted in).

Regards
Mohsen

Thanks so much Mohsen,

Do you know how I can stop the plate (if I follow the Odeometric approach ) when I reach to a specific porosity. I don’t use triaxialStressController in this case. I can compact the layer but I can’t stop the plate by target porosity criterion.

If we resolve this, I need to know how to save and load the final position in my main triaxial script.

Is it possible to compact layers by triaxial goal in one direction while all other walls are deactivated?

Sincerely,

Nima

> On Oct 8, 2017, at 5:28 AM, mohsen <email address hidden> wrote:
>
> Your question #658905 on Yade changed:
> https://answers.launchpad.net/yade/+question/658905
>
> Status: Open => Answered
>
> mohsen proposed the following answer:
>>> Thanks again for your hint. Do you mean I need to have two distinct
> scripts if I want to follow the instructions like in odeometer test?
>
> Nima If you follow this, you need to scripts. the first one is for
> sample generation and compaction; and the second one for shearing. i do
> not have any script regarding this approach.
>
>>> I don't need an exact script for what I am doing, even if you have
> an approximate solution (script) regarding deactivating walls that would
> be enough for me.
>
> You can easily deactivates walls:
>
> wall_front_activated=False
>
> there are 6 walls which in your case all of them should be deactivate.
> you may define a plate for compaction. After compaction of each layer
> you may move or delete the plate and then you can generate again the
> next layer and compact and so on
>
>
>> Another critical question is about how recognizing the surface of previous layers. I mean, I am compacting layers based on the geometry before compaction. How we can enforce the code to understand the new position of layers (top surface of layers will be lower after compaction but I am generating the next layer exactly from the original height of top surface of previous layer (before compaction)). Is there any solution?
>
> You easily can distinguish the new position of the recent layer that has
> been just compacted:
>
> MaxZ=max([b.state.pos[2]+b.shape.radius for b in O.bodies if isinstance(b.shape,Sphere)])
> MaxZ is the coordination of the grain with maximum height (Note that Z is defined as the axis which grain are compacted in).
>
> Regards
> Mohsen
>
> --
> If this answers your question, please go to the following page to let us
> know that it is solved:
> https://answers.launchpad.net/yade/+question/658905/+confirm?answer_id=4
>
> If you still need help, you can reply to this email or go to the
> following page to enter your feedback:
> https://answers.launchpad.net/yade/+question/658905
>
> You received this question notification because you asked the question.

>>Do you know how I can stop the plate (if I follow the Odeometric approach ) when I reach to a specific porosity. I don’t use triaxialStressController in this case. I can compact the layer but I can’t stop the plate by target porosity criterion.

you can use:

        Por_osity=voxelPorosity(200,mn,mx)
        if Por_osity<targetPorosity:
            plate.state.vel=(0,0,0)

Be careful you should define mn and mx for each layer specifically. maxz would help to define mx:
        maxz=max([b.state.pos[2]+b.shape.radius for b in O.bodies if isinstance(b.shape,Sphere)])

>>If we resolve this, I need to know how to save and load the final position in my main triaxial script:
Saving:
# creating a list with position and radius of particles
ParticlePos=[]
for b in O.bodies:
    if isinstance(b.shape,Sphere):
        ParticlePos.append([b.state.pos,b.shape.radius])

#saving the position and radius of particles
with open(str(ParticlePos.txt","wb") as f:
    pickle.dump(ParticlePos,f)

Nw Then you have a text file with position and radius of particles
Loading:
ParticlePos=[]
with open("ParticlePos.txt","rb") as f:
    ParticlePos=pickle.load(f)

Generating spheres:

while i+1<len(ParticlePos):
    # here you can optomize your grid network
    i=i+1
    c,r=ParticlePos[i]
    s = utils.sphere(c,r,material=labelS)
    O.bodies.append(s)

>>Is it possible to compact layers by triaxial goal in one direction while all other walls are deactivated?:
I think defining a plate is easier since you need to compact several layers. But you can compact the total specimen by deactivating all of wall except the desired one.

Mohsen

 Hi Mohsen,
Thanks sincerely for your hint. I tried to implement what you said for odeometric-like approach with no success. I'm bringing this part of my code. If possible please take a look to see if you can fix it.
from yade.pack import *from yade import utilsfrom yade import exportfrom yade import plotn_s = 2000n_band = 26# For now I'm using Sobieski's PSD. I need to change it to my own PSD latertargetPorosity= 0.422##################################################   Particle Size Distribution   ##################################################psdSizes=[5.8598400e+00,5.8761600e+00,5.8924800e+00,5.9088000e+00,5.9251200e+00,5.9414400e+00,5.9577600e+00,5.9740800e+00,5.9904000e+00,6.0067200e+00,6.0230400e+00,6.0393600e+00,6.0556800e+00,6.0720000e+00,6.0883200e+00,6.1046400e+00,6.1209600e+00,6.1372800e+00,6.1536000e+00,6.1699200e+00,6.1862400e+00,6.2025600e+00,6.2188800e+00,6.2352000e+00,6.2515200e+00,6.2678400e+00]psdCumm=[0.0,8.4945735e-05,3.5804112e-04,1.1512195e-03,3.2324246e-03,8.1658647e-03,1.8731095e-02,3.9172232e-02,7.4902028e-02,1.3132521e-01,2.1182373e-01,3.1558203e-01,4.3640887e-01,5.6352779e-01,6.8435463e-01,7.8811293e-01,8.6861145e-01,9.2503463e-01,9.6076443e-01,9.8120556e-01,9.9177079e-01,9.9670423e-01,9.9878544e-01,9.9957862e-01,9.9985171e-01,1.0]
i = 0while i < n_band:    psdSizes[i]=psdSizes[i]/1000. #scaling from [mm] to [m]    print "\r psdSizes:",i,psdSizes[i]    i = i +1
#If the bed volume is defined, then the average diameter of particles (when the target porosity is reached) is the same like in the measurement###########################################################   facetBox   ############################################################ Compaction will be in Y-Directionl = 3.5e-3h = 7.1e-3vol = l*h*l# You maybe don't need this if you use compaction from the top of the sample (using plate)
mn = Vector3(0,0,0)mx = Vector3(l,l,h/5)
O.materials.append(FrictMat(young=5e6,poisson=0.5,frictionAngle=0,density=0,label='walls'))O.materials.append(FrictMat(young=5e6,poisson=0.5,frictionAngle=radians(30),density=2600,label='spheres'))O.bodies.append(geom.facetBox((l/2,l/2,h/2),(l/2,l/2,h/2),wallMask=31,material='walls'))# This will be an open-top boxsp = yade._packSpheres.SpherePack()sp.makeCloud(mn,mx,psdSizes=psdSizes,psdCumm=psdCumm,num=n_s,distributeMass=1)O.bodies.append([sphere(s[0],s[1],material='spheres') for s in sp])Porosity = voxelPorosity(200,mn,mx)print "\r Porosity: ",Porosity,O.engines=[   ForceResetter(),   # sphere, facet, wall   InsertionSortCollider([Bo1_Sphere_Aabb(),Bo1_Facet_Aabb(),Bo1_Wall_Aabb()]),   InteractionLoop(      # the loading plate is a wall, we need to handle sphere+sphere, sphere+facet, sphere+wall      [Ig2_Sphere_Sphere_ScGeom(),Ig2_Facet_Sphere_ScGeom(),Ig2_Wall_Sphere_ScGeom()],      [Ip2_FrictMat_FrictMat_FrictPhys()],      [Law2_ScGeom_FrictPhys_CundallStrack()]   ),   NewtonIntegrator(gravity=(0,0,-9.81),damping=0.5),   # the label creates an automatic variable referring to this engine   # we use it below to change its attributes from the functions called   PyRunner(command='checkUnbalanced()',realPeriod=2,label='checker'),]O.dt=.5*PWaveTimeStep()
def checkUnbalanced():   # at the very start, unbalanced force can be low as there is only few contacts, but it does not mean the packing is stable   if O.iter<5000: return    # the rest will be run only if unbalanced is < .1 (stabilized packing)   if unbalancedForce()>.1: return    # add plate at the position on the top of the packing   # the maximum finds the z-coordinate of the top of the topmost particle   O.bodies.append(wall(max([b.state.pos[2]+b.shape.radius for b in O.bodies if isinstance(b.shape,Sphere)]),axis=2,sense=-1))   global plate        # without this line, the plate variable would only exist inside this function   plate=O.bodies[-1]  # the last particles is the plate   # Wall objects are "fixed" by default, i.e. not subject to forces   # prescribing a velocity will therefore make it move at constant velocity (downwards)   plate.state.vel=(0,0,-.1)   # start plotting the data now, it was not interesting before   O.engines=O.engines+[PyRunner(command='addPlotData()',iterPeriod=200)]   # next time, do not call this function anymore, but the next one (unloadPlate) instead   checker.command='unloadPlate()'def unloadPlate():   # if the force on plate exceeds maximum load, start unloading   maxz=max([b.state.pos[2]+b.shape.radius for b in O.bodies if isinstance(b.shape,Sphere)])   mx=maxz# Iknow this doesn't work   Por_osity=voxelPorosity(200,mn,mx)   if Por_osity <targetPorosity:      plate.state.vel=(0,0,0)
Thanks so much,
Nima

    On Tuesday, October 10, 2017, 1:04:19 AM CDT, mohsen <email address hidden> wrote:

 Your question #658905 on Yade changed:
https://answers.launchpad.net/yade/+question/658905

    Status: Open => Answered

mohsen proposed the following answer:
>>Do you know how I can stop the plate (if I follow the Odeometric
approach ) when I reach to a specific porosity. I don’t use
triaxialStressController in this case. I can compact the layer but I
can’t stop the plate by target porosity criterion.

you can use:

        Por_osity=voxelPorosity(200,mn,mx)
        if Por_osity<targetPorosity:
            plate.state.vel=(0,0,0)

Be careful you should define mn and mx for each layer specifically. maxz would help to define mx:
        maxz=max([b.state.pos[2]+b.shape.radius for b in O.bodies if isinstance(b.shape,Sphere)])

>>If we resolve this, I need to know how to save and load the final position in my main triaxial script:
Saving:
# creating a list with position and radius of particles
ParticlePos=[]
for b in O.bodies:
    if isinstance(b.shape,Sphere):
        ParticlePos.append([b.state.pos,b.shape.radius])

#saving the position and radius of particles
with open(str(ParticlePos.txt","wb") as f:
    pickle.dump(ParticlePos,f)

Nw Then you have a text file with position and radius of particles
Loading:
ParticlePos=[]
with open("ParticlePos.txt","rb") as f:
    ParticlePos=pickle.load(f)

Generating spheres:

while i+1<len(ParticlePos):
    # here you can optomize your grid network
    i=i+1
    c,r=ParticlePos[i]
    s = utils.sphere(c,r,material=labelS)
    O.bodies.append(s)

>>Is it possible to compact layers by triaxial goal in one direction while all other walls are deactivated?:
I think defining a plate is easier since you need to compact several layers. But you can compact the total specimen by deactivating all of wall except the desired one.

Mohsen

--
If this answers your question, please go to the following page to let us
know that it is solved:
https://answers.launchpad.net/yade/+question/658905/+confirm?answer_id=6

If you still need help, you can reply to this email or go to the
following page to enter your feedback:
https://answers.launchpad.net/yade/+question/658905

You received this question notification because you asked the question.

 Sorry mohsen. Attached please find the code. Please ignore the one I sent you in the last email.
Thanks
Nima

    On Tuesday, October 17, 2017, 5:10:23 PM CDT, Nima Goudarzi <email address hidden> wrote:

 Your question #658905 on Yade changed:
https://answers.launchpad.net/yade/+question/658905

    Status: Answered => Open

You are still having a problem:
 Hi Mohsen,
Thanks sincerely for your hint. I tried to implement what you said for odeometric-like approach with no success. I'm bringing this part of my code. If possible please take a look to see if you can fix it.
from yade.pack import *from yade import utilsfrom yade import exportfrom yade import plotn_s = 2000n_band = 26# For now I'm using Sobieski's PSD. I need to change it to my own PSD latertargetPorosity= 0.422##################################################   Particle Size Distribution   ##################################################psdSizes=[5.8598400e+00,5.8761600e+00,5.8924800e+00,5.9088000e+00,5.9251200e+00,5.9414400e+00,5.9577600e+00,5.9740800e+00,5.9904000e+00,6.0067200e+00,6.0230400e+00,6.0393600e+00,6.0556800e+00,6.0720000e+00,6.0883200e+00,6.1046400e+00,6.1209600e+00,6.1372800e+00,6.1536000e+00,6.1699200e+00,6.1862400e+00,6.2025600e+00,6.2188800e+00,6.2352000e+00,6.2515200e+00,6.2678400e+00]psdCumm=[0.0,8.4945735e-05,3.5804112e-04,1.1512195e-03,3.2324246e-03,8.1658647e-03,1.8731095e-02,3.9172232e-02,7.4902028e-02,1.3132521e-01,2.1182373e-01,3.1558203e-01,4.3640887e-01,5.6352779e-01,6.8435463e-01,7.8811293e-01,8.6861145e-01,9.2503463e-01,9.6076443e-01,9.8120556e-01,9.9177079e-01,9.9670423e-01,9.9878544e-01,9.9957862e-01,9.9985171e-01,1.0]
i = 0while i < n_band:    psdSizes[i]=psdSizes[i]/1000. #scaling from [mm] to [m]    print "\r psdSizes:",i,psdSizes[i]    i = i +1
#If the bed volume is defined, then the average diameter of particles (when the target porosity is reached) is the same like in the measurement###########################################################   facetBox   ############################################################ Compaction will be in Y-Directionl = 3.5e-3h = 7.1e-3vol = l*h*l# You maybe don't need this if you use compaction from the top of the sample (using plate)
mn = Vector3(0,0,0)mx = Vector3(l,l,h/5)
O.materials.append(FrictMat(young=5e6,poisson=0.5,frictionAngle=0,density=0,label='walls'))O.materials.append(FrictMat(young=5e6,poisson=0.5,frictionAngle=radians(30),density=2600,label='spheres'))O.bodies.append(geom.facetBox((l/2,l/2,h/2),(l/2,l/2,h/2),wallMask=31,material='walls'))# This will be an open-top boxsp = yade._packSpheres.SpherePack()sp.makeCloud(mn,mx,psdSizes=psdSizes,psdCumm=psdCumm,num=n_s,distributeMass=1)O.bodies.append([sphere(s[0],s[1],material='spheres') for s in sp])Porosity = voxelPorosity(200,mn,mx)print "\r Porosity: ",Porosity,O.engines=[   ForceResetter(),   # sphere, facet, wall   InsertionSortCollider([Bo1_Sphere_Aabb(),Bo1_Facet_Aabb(),Bo1_Wall_Aabb()]),   InteractionLoop(      # the loading plate is a wall, we need to handle sphere+sphere, sphere+facet, sphere+wall      [Ig2_Sphere_Sphere_ScGeom(),Ig2_Facet_Sphere_ScGeom(),Ig2_Wall_Sphere_ScGeom()],      [Ip2_FrictMat_FrictMat_FrictPhys()],      [Law2_ScGeom_FrictPhys_CundallStrack()]   ),   NewtonIntegrator(gravity=(0,0,-9.81),damping=0.5),   # the label creates an automatic variable referring to this engine   # we use it below to change its attributes from the functions called   PyRunner(command='checkUnbalanced()',realPeriod=2,label='checker'),]O.dt=.5*PWaveTimeStep()
def checkUnbalanced():   # at the very start, unbalanced force can be low as there is only few contacts, but it does not mean the packing is stable   if O.iter<5000: return    # the rest will be run only if unbalanced is < .1 (stabilized packing)   if unbalancedForce()>.1: return    # add plate at the position on the top of the packing   # the maximum finds the z-coordinate of the top of the topmost particle   O.bodies.append(wall(max([b.state.pos[2]+b.shape.radius for b in O.bodies if isinstance(b.shape,Sphere)]),axis=2,sense=-1))   global plate        # without this line, the plate variable would only exist inside this function   plate=O.bodies[-1]  # the last particles is the plate   # Wall objects are "fixed" by default, i.e. not subject to forces   # prescribing a velocity will therefore make it move at constant velocity (downwards)   plate.state.vel=(0,0,-.1)   # start plotting the data now, it was not interesting before   O.engines=O.engines+[PyRunner(command='addPlotData()',iterPeriod=200)]   # next time, do not call this function anymore, but the next one (unloadPlate) instead   checker.command='unloadPlate()'def unloadPlate():   # if the force on plate exceeds maximum load, start unloading   maxz=max([b.state.pos[2]+b.shape.radius for b in O.bodies if isinstance(b.shape,Sphere)])   mx=maxz# Iknow this doesn't work   Por_osity=voxelPorosity(200,mn,mx)   if Por_osity <targetPorosity:      plate.state.vel=(0,0,0)
Thanks so much,
Nima

    On Tuesday, October 10, 2017, 1:04:19 AM CDT, mohsen <email address hidden> wrote:

 Your question #658905 on Yade changed:
https://answers.launchpad.net/yade/+question/658905

    Status: Open => Answered

mohsen proposed the following answer:
>>Do you know how I can stop the plate (if I follow the Odeometric
approach ) when I reach to a specific porosity. I don’t use
triaxialStressController in this case. I can compact the layer but I
can’t stop the plate by target porosity criterion.

you can use:

        Por_osity=voxelPorosity(200,mn,mx)
        if Por_osity<targetPorosity:
            plate.state.vel=(0,0,0)

Be careful you should define mn and mx for each layer specifically. maxz would help to define mx:
        maxz=max([b.state.pos[2]+b.shape.radius for b in O.bodies if isinstance(b.shape,Sphere)])

>>If we resolve this, I need to know how to save and load the final position in my main triaxial script:
Saving:
# creating a list with position and radius of particles
ParticlePos=[]
for b in O.bodies:
    if isinstance(b.shape,Sphere):
        ParticlePos.append([b.state.pos,b.shape.radius])

#saving the position and radius of particles
with open(str(ParticlePos.txt","wb") as f:
    pickle.dump(ParticlePos,f)

Nw Then you have a text file with position and radius of particles
Loading:
ParticlePos=[]
with open("ParticlePos.txt","rb") as f:
    ParticlePos=pickle.load(f)

Generating spheres:

while i+1<len(ParticlePos):
    # here you can optomize your grid network
    i=i+1
    c,r=ParticlePos[i]
    s = utils.sphere(c,r,material=labelS)
    O.bodies.append(s)

>>Is it possible to compact layers by triaxial goal in one direction while all other walls are deactivated?:
I think defining a plate is easier since you need to compact several layers. But you can compact the total specimen by deactivating all of wall except the desired one.

Mohsen

--
If this answers your question, please go to the following page to let us
know that it is solved:
https://answers.launchpad.net/yade/+question/658905/+confirm?answer_id=6

If you still need help, you can reply to this email or go to the
following page to enter your feedback:
https://answers.launchpad.net/yade/+question/658905

You received this question notification because you asked the question.

You received this question notification because you asked the question.

This question was expired because it remained in the 'Open' state without activity for the last 15 days.