Hello，I have ran the simulation with the in.noGB_small.However，it had a error after runing 200000 steps .
Error：Illegal fix_modify command! Can you please tell me why it happend？
In additional，I found a new command “fix_modify AtC transfer lumped_lambda_solve on” in the input， what does it ？The command is not exist in the Atc package.
Thank you very much!
Best regards！
This sounds like a Q for the AtC folks.
Steve
Ah, sorry, you'll note that the localization thermostat was developed for Templeton et al, but the current atc release is based on Wagner et al. This is good impetus for us to clean up our current version and push it out. For now, commenting out the line should work.
Jeremy
Hello,thank you very much! I’m looking forward to the new version.However,I have the following questions about the in.noGB_small. I need your help.
Q1**:In the y direction ,the extent is (0 7.68052) in the region “simregion” while the extent is (0 9.1337324003) in the simulation box created by the command read_date. Why is the extent in the region “simregion” different from the box ?In fact , I think that they can be the same without any effect.
Q2:In the z direcion, the extent is (5.48722 272.7937514) in the region “simregion” while the extent is (0 282.2004284896) in the simulation box.Are the ghost atoms in the (0 5.48722)and (272.7937514 282.2004284896)? I find that the extent in the ghost region is less than 3a(a=lattice spacing), however, the extent for ghost atoms is about 4a in the paper (Templeton et al).Why are they different ?
Q3:What is the paramfile layout in the command "fix ID groupID atc type paramfile ".Why did you give the diffusivity(not the other quantitys) in the file？what is the meaning of diffusivity=0.0 in the file？
Q4: I find that the command “fix_modify AtC transfer boundary type ghost” is not exist in the input. Why ？
Q5:What is the meaning of the command “fix_modify AtC transfer internal_quadrature off”?Can you please tell me how to choice "on"or “off”?
Q6:Why did not remove the contributions of ghost atoms in the temp by the command “variable xdof equal 3*count(ghost) compute_modify thermo_temp extra ${xdof}” ？
Q7:The syntax is “fix_modify AtC create_faceset <in|out> [units]”. However, I find that the in(or out) and the units keyword are not specified in the input . Why was not the in(or out) be specified? Though I know the default is unit=lattice for units keyword, I think that the unit=box should be specified for the units keyword in the in.noGB_small.Is it right ?
Q8:What is the meaning of the command “fix_modify AtC transfer thermal control flux no_boundary” ? In fact.the syntax is “fix_modify AtC transfer thermal control flux <boundary_integration_type> <face_set_id(optional)>”, but the <boundary_integration_type> is not specified in the in.noGB_small**.What is the meaning of the no_boundary ?
Q9:You have told me that the heat flux is applied to the MD boundary in the example 5.1 problem.However, I did’t find the command which applied the heat flux to MD boundary. By contrast, I only find the command “fix_modify AtC transfer fix_flux temperature lbc 0.00000001” which applied the heat flxu to FE boundary .
Q10:In the example 5.1 , the heat flux is 3.32*(10^9)(w/m/m) .However. the 0.00000001 in real units is corresponding to 6.9535*(10^6)(w/m/m) which is different from 3.32*(10^9).Why?
Q11:The “time filter” was not used in the input.Why? Can you please tell me how to determine if the time filter is used?
Q12:Why is the epsilon in Si.sw different from the epsilon in the Si.sw of the potential directory in lammps ? In fact ,I have done the unit conversions.
Thank you very much and look forward to hearing from you soon.
Best regards!
2011/9/17 Templeton, Jeremy Alan <jatempl@…3…>
Let me emphasize that you should read the on-line docs and the talk from the last Lammps users workshop. Many of the questions are discussed on more detail there.
Q1**:**In the y direction ,the extent is (0 7.68052) in the region “simregion” while the extent is (0 9.1337324003) in the simulation box created by the command read_date. Why is the extent in the region “simregion” different from the box ?In fact , I think that they can be the same without any effect.
These were the dimensions I had to put into lammps to get the lattice with the extent I wanted. Others are more knowledgeable than I in creating lammps.
Q2:In the z direcion, the extent is (5.48722 272.7937514) in the region “simregion” while the extent is (0 282.2004284896) in the simulation box.Are the ghost atoms in the (0 5.48722)and (272.7937514 282.2004284896)? I find that the extent in the ghost region is less than 3a(a=lattice spacing), however, the extent for ghost atoms is about 4a in the paper (Templeton et al).Why are they different ?
Again, the boxes are only relevant in terms of the FE mesh. Other than that, they don’t matter.
Q3:What is the paramfile layout in **the command "**fix ID groupID atc type paramfile ".Why did you give the diffusivity(not the other quantitys) in the file？what is the meaning of diffusivity=0.0 in the file？
See the documentation. For boundary condition simulations, the diffusivity is irrelevant (see also the talk).
Q4: I find that the command “fix_modify AtC transfer boundary type ghost” is not exist in the input. Why ？
Again, see the talk. Ghost atoms are only needed to define the geometry in multiscale simulations
Q5:What is the meaning of the command “fix_modify AtC transfer internal_quadrature off”?Can you please tell me how to choice "on"or “off”?
See the documentation (but you want off).
Q6:Why did not remove the contributions of ghost atoms in the temp by the command “variable xdof equal 3*count(ghost) compute_modify thermo_temp extra ${xdof}” ？
Because it wasn’t relevant for what I wanted.
Q7:The syntax is “fix_modify AtC create_faceset <in|out> [units]”. However, I find that the in(or out) and the units keyword are not specified in the input . Why was not the in(or out) be specified? Though I know the default is unit=lattice for units keyword, I think that the unit=box should be specified for the units keyword in the in.noGB_sm**all.Is it right ?**
This command is described in the documentation.
Q8:What is the meaning of the command “fix_modify AtC transfer thermal control flux no_boundary” ? In fact.the syntax is “fix_modify AtC transfer thermal control flux <boundary_integration_type> <face_set_id(optional)>”, but the <boundary_integration_type> is not specified in the in.noGB_small.What is the meaning of the no_boundary ?
Sorry, this has not yet made it to the documentation. No boundary means there is multiscale heat flux.
Q9:You have told me that the heat flux is applied to the MD boundary in the example 5.1 problem.However, I did’t find the command which applied the heat flux to MD boundary. By contrast, I only find the command “fix_modify AtC transfer fix_flux temperature lbc 0.00000001” which applied the heat flxu to FE boundary .
Applying a heat flux is precisely what this command does.
Q10:In the example 5.1 , the heat flux is 3.32*(10^9)(w/m/m) .However. the 0.00000001 in real units is corresponding to 6.9535*(10^6)(w/m/m) which is different from 3.32*(10^9).Why?
Either I’ve modified the file and forgotten or you made a factor of two error. We compared our results with theory in the paper so I have confidence in our results.
Q11:The “time filter” was not used in the input.Why? Can you please tell me how to determine if the time filter is used?
It’s entirely up to you, but it was developed in Wagner et al. so that the molecular temperature was more consistent with a continuum model (Fourier’s law).
Q12:Why is the epsilon in Si.sw different from the epsilon in the Si.sw of the potential directory in lammps ? In fact ,I have done the unit conversions.
This is a set of parameters a collaborator (S. Aubry) suggested. Use whatever parameters you like.
Hello,Thank you very much for your help! I’m very grateful. I have done my best to read the on-line docs and the talk .However,I still can’t understand some questions mentioned.Please help me !
Q6:Why did not remove the contributions of ghost atoms in the “ temp" by the command “variable xdof equal 3*count(ghost) compute_modify thermo_temp extra ${xdof}” ？What is the temp’s significance ? In other words ,If I want to get the atomic temperature ,What should I do ?
Q7:
(1).I have read the on-line docs(http://lammps.sandia.gov/doc/fix_atc.html) and the talk. The syntax is “fix_modify AtC create_faceset <in|out> [units]”. However, I find that the in(or out) keyword is not specified in the in.noGB_small.The commands are “fix_modify AtC mesh create_faceset lbc -20 20 -20 20 5.487228542885853 300"and “fix_modify AtC mesh create_faceset rbc -20 20 -20 20 0 272.793751436214147”.The docs does’t tell us what is the default value for in/out keyword.However,I have done my best to read the source code about the command and I think the “outward” is the default value. Is it right?
(2).In additional, I find a new question about the command. I really can’t understand it.In the soure code ,the command have a ”box/plane“ keyword which is not mentioned in the docs .What do the “box"and"plane” do,respectively? I think that the default value is the “box” keyword by reading the soure code.Is it right?
(3).If the default value is “box”, I have tested the in_noGBsmall.First, the input ran with the “fix_modify AtC mesh create_faceset lbc -20 20 -20 20 5.487228542885853 300”. The screen showed " ATC:: created faceset lbc with 1 faces” . Then, the input ran with the “fix_modify AtC mesh create_faceset lbc box -20 20 -20 20 5.487228542885853 300”.The screen showed “ATC:: created faceset lbc with 11 faces”. Why ? In fact, I am very puzzled about the command . Can you please tell me how to know which faces will be created or Can you please tell me the algorithm about creating the faceset?
Q10:I have calculated the heat flux many times .However, the 0.00000001 in real units still is corresponding to 6.9535*(10^6)(w/m/m) which is different from 3.32*(10^9) .My process : 0. 00000001=10^(-8)(Kcal/mol/fs/A/A)=10^(-8)4186/(6.0210^(23))/10^(-15)/10^(-20)(w/m/m)=6.9535*(10^6)(w/m/m).Is it right ? If it isn’t, can you please tell me how to transition it?
A new question below:
Q1: How to apply a temperature profile to FE boundary? I have read the talk and I only found the command "fix_modify AtC fix temperature lbc temporal_ramp 20 40 25000 " on page 20 . Is the command used to apply the temperature profile ? If it is ,what are the meaning of "20 、40 、25000”，respectively？
Thanks again !
Best regards!
wangdan
在 2011年9月20日 上午11:31，Templeton, Jeremy Alan <jatempl@…3…>写道：
Hello Wangdan, answers below:
Q6:Why did not remove the contributions of ghost atoms in the “ temp" by the command “variable xdof equal 3*count(ghost) compute_modify thermo_temp extra ${xdof}” ？What is the temp’s significance ? In other words ,If I want to get the atomic temperature ,What should I do ?
As I said, I didn’t do it because I didn’t need that data. If you need it, you can add it. The input decks in the release are designed to show off atc capabilities, not lammps capabilities. It is our hope that users such as yourself will use their knowledge of lammps combined with the atc examples to create problems that fit your needs. The paper example I gave you is even worse in that regard because I never expected anybody but me to look at it.
Q7:
(1).I have read the on-line docs(http://lammps.sandia.gov/doc/fix_atc.html) and the talk. The syntax is “fix_modify AtC create_faceset <in|out> [units]”. However, I find that the in(or out) keyword is not specified in the in.noGB_small.The commands are "fix_modify AtC mesh create_faceset lbc -20 20 -20 20 5.487228542885853 300"and “fix_modify AtC mesh create_faceset rbc -20 20 -20 20 0 272.793751436214147”.The docs does’t tell us what is the default value for in/out keyword.However,I have done my best to read the source code about the command and I think the “outward” is the default value. Is it right?
My apologies, the in/out default is not properly documented. The default is out.
(2).In additional, I find a new question about the command. I really can’t understand it.In the soure code ,the command have a ”box/plane“ keyword which is not mentioned in the docs .What do the “box"and"plane” do,respectively? I think that the default value is the “box” keyword by reading the soure code.Is it right?
The “plane” feature is not something we’re supporting yet (and hence why it’s not documented), but we added for our work. Use at your own risk.
(3).If the default value is “box”, I have tested the in_noGBsmall.First, the input ran with the “fix_modify AtC mesh create_faceset lbc -20 20 -20 20 5.487228542885853 300”. The screen showed " ATC:: created faceset lbc with 1 faces" . Then, the input ran with the “fix_modify AtC mesh create_faceset lbc box -20 20 -20 20 5.487228542885853 300”.The screen showed “ATC:: created faceset lbc with 11 faces”. Why ? In fact, I am very puzzled about the command . Can you please tell me how to know which faces will be created or Can you please tell me the algorithm about creating the faceset?
When I try both versions of the command, I get the same result, so I’m not sure why you don’t. The correct value should be 1 faceset is created and in general I suggest sticking with the syntax described in the documentation. The ‘box’ mode creates a box using the dimensions you specify, and then creates a faceset of all the element faces that intersect the boundary of that block.
Q10:I have calculated the heat flux many times .However, the 0.00000001 in real units still is corresponding to 6.9535*(10^6)(w/m/m) which is different from 3.32*(10^9) .My process : 0. 00000001=10^(-8)(Kcal/mol/fs/A/A)=10^(-8)4186/(6.0210^(23))/10^(-15)/10^(-20)(w/m/m)=6.9535*(10^6)(w/m/m).Is it right ? If it isn’t, can you please tell me how to transition it?
As noted on the doc page (http://lammps.sandia.gov/doc/USER/atc/man_fix_flux.html): “The units are in AtC units, i.e. derived from the LAMMPS length, time, and mass scales.” That should point you to the discrepancy.
A new question below:
Q1: How to apply a temperature profile to FE boundary? I have read the talk and I only found the command "fix_modify AtC fix temperature lbc temporal_ramp 20 40 25000 " on page 20 . Is the command used to apply the temperature profile ? If it is ,what are the meaning of "20 、40 、25000”，respectively？
This can be done if you modify XT_Function.h/cpp to have the functional form you require. If you add some useful functions and are willing to share, you can send the modified code to us for inclusion in our next release.
Hello,I‘m very grateful for your help.
In Q7 (3), I am very sorry for my fault about those two different results.I have understood the command with your help.
However, I still can’t understand the question 10.In fact,I have read the sentence “The units are in AtC units, i.e. derived from the LAMMPS length, time, and mass scales” before asking you the question.So, If you don’t mind ,Can you please write the detailed process about transforming the 0. 00000001 to 3.32*(10^9)(w/m/m) ? I really don’t know what’s worry with my idea.
Thanks again !
Best regards!
wangdan
在 2011年9月23日 上午1:13，Templeton, Jeremy Alan <jatempl@…3…>写道：