hi, ATC workers
I successfully solved some problems with your help. In order to test if my code are correct, I chose the work of Zhigilei in “Combined atomistic-continum modeling of short-pulse laser melting and disintegration of metal films” to repeat (TTM model). But the result is so different , the temperature of mine is lower than it in Zhigilei’s work. in my mind ,since your development built upon the work of Zhigilei and co-workers, the result should be similar.
Q1: I think the main problem is the way of applying laser energy. I use" fix_modify AtC fix_flux electron_temperature bndy**" to apply laser energy , but in TTM equations ,the source term represent energy per volume per time,i.e. ,source not flux in AtC. so I need to change the laser fluence to S(z,t) and use "fix_modify AtC source electron_temperature ", am I correct?
Q2: A Gaussian temporal profile is used in the simulations to describe the laser energy deposition , how to realize in AtC?
Q3: At last ,I want to get contour plots of lattice temperature, so I should to output temperature
Hi Emily, I’m glad you are making progress. Below are some suggestions:
Q1: I think the main problem is the way of applying laser energy. I use" fix_modify AtC fix_flux electron_temperature bndy**" to apply laser energy , but in TTM equations ,the source term represent energy per volume per time,i.e. ,source not flux in AtC. so I need to change the laser fluence to S(z,t) and use "fix_modify AtC source electron_temperature ", am I correct?
In the end, all energy sources in AtC are applied volumetrically, but the way it is specified is important to know what the volumetric heating is, as well as its spatial extent. I suggest you figure out the rate of energy you want deposited and then convert that into a surface flux or volumetric source as you like. Certainly taking a volumetric source and using it as a surface flux would not provide the correct answer.
Q2: A Gaussian temporal profile is used in the simulations to describe the laser energy deposition , how to realize in AtC?
See the examples: two_temperature/in.gaussianIC_ttm and two_temperature/in.no_atoms to see how to specify a Gaussian function. The files Function.h/cpp are there to allow users to specify their own type of function, so you will find more information on the Gaussian function there.
Q3: At last ,I want to get contour plots of lattice temperature, so I should to output temperature
Yes.
dear Jeremy
in your last mail,you said “In the end, all energy sources in AtC are applied volumetrically…”, do you mean as long as I convert my rate of energy deposited into a surface flux or volumetric source correctly, both of the applying ways are OK? they won’t cause different results?
Q1: if I choose surface flux to describe my short-pulse laser energy deposition (uniform distribution in space,Gaussian distribution in time and an exponential attenuation of laser intensity with depth under surface,i.e. Beer-Lambert law), how to reflect the optical absorption depth Lp and the reflectivity R?
Q2 If I want to get contour plots of lattice temperature,in my mind ,I shoud output temperature of each node at each time,but the AtC output are globle values,so I achieve my goal through commands below:
compute ke all ke/atom
variable temp atom c_ke/1.5/8.617343e-5
variable z atom z
fix 3 all ave/spatial 10 1 10 z upper 10 v_z v_temp units box file T.profile
but there are some incredible temperature data in the file like 6.90157K or 1.55295e+06K.do you have some better advise to achieve my goal?
Q3 can you explain the true meaning of the faceset? in my mind , the MD surface should be consistent with the faceset ,and we apply laser energy flux through it, in that case, why the faceset should be describe as a box?
Emily
Emily,
in your last mail,you said “In the end, all energy sources in AtC are applied volumetrically…”, do you mean as long as I convert my rate of energy deposited into a surface flux or volumetric source correctly, both of the applying ways are OK? they won’t cause different results?
The volumetric application of fluxes is part of the AtC theory. I suggest you take some time and go over the paper which describes how volumetric and surface sources are incorporated. You can specify sources in multiple ways, but need to be consistent with what the physics are and what the method needs.
Q1: if I choose surface flux to describe my short-pulse laser energy deposition (uniform distribution in space,Gaussian distribution in time and an exponential attenuation of laser intensity with depth under surface,i.e. Beer-Lambert law), how to reflect the optical absorption depth Lp and the reflectivity R?
That is part of your research to figure this out. Again, the paper will spell out what is there, but it is the job of the user to derive complex physics laws as the domain expert. AtC allows you to add material models and we can point you where in the code to do that when you have the model ready.
Q2 If I want to get contour plots of lattice temperature,in my mind ,I shoud output temperature of each node at each time,but the AtC output are globle values,so I achieve my goal through commands below:
compute ke all ke/atom
variable temp atom c_ke/1.5/8.617343e-5
variable z atom z
fix 3 all ave/spatial 10 1 10 z upper 10 v_z v_temp units box file T.profilebut there are some incredible temperature data in the file like 6.90157K or 1.55295e+06K.do you have some better advise to achieve my goal?
The documentation provides examples for outputting nodal variables (http://lammps.sandia.gov/doc/USER/atc/man_output.html).
Q3 can you explain the true meaning of the faceset? in my mind , the MD surface should be consistent with the faceset ,and we apply laser energy flux through it, in that case, why the faceset should be describe as a box?
A face set is a collection of finite element surfaces (the reference by Hughes will cover these topics). One way we allow the user to create them is by intersecting faces with a prescribed box, so only the faces which match its surface are included. There is an AtC command which will output the mesh including face sets and node sets for your to check: http://lammps.sandia.gov/doc/USER/atc/man_mesh_output.html