laser ablation,fix atc

Dear everyone
I want to use atc package to simulate femtosecond laser ablation in Cu.My goal is analysing the ablation mechnical through getting snap shots and temperature evolution(distribution) and so on.Because the pulse duration of my simulation is too short(100fs),so I use two-temperature model(TTM) to simulate the energy translation. In my simulation,laser is just a heat source.And I assume the energy is adding average in spatial and time .Here is my code and question:
#initialization
units metal
variable inf equal 1000
dimension 3
boundary p p s
Q1: add laser energy along z coordinate,and I want to trace every atoms in z. is the boundary condition correct?

#atom definition
lattice fcc 3.62
region MD block 0 5 0 5 50 250
region FE block 0 5 0 5 0 250
create_box 1 MD
create_atoms 1 region MD
#set potential
pair_style eam
pair_coeff * * Cu.eam
neighbor 0.2 bin
neigh_modify every 10 delay 0 check yes

dump 1 all atom 100 dump.laser
group internal region MD

timestep 0.001
thermo 200

velocity internal create 300 1234156
fix 1 internal nve
run 1000
unfix 1

#ttm-md coupling
fix AtC internal atc two_temperature Cu_ttm.mat
fix_modify AtC mesh create 1 1 80 FE p p f

fix_modify AtC mesh create_faceset bndy box -inf inf -inf inf 50 250
Q2:as I know, applying an incident flux should be done through a faceset.and I found the faceset can affect the final result.If I set the faceset region as “box 0 5 0 5 50 250”,the final temperature will become too high and not convergence. To my mind the region"-inf inf -inf inf 50 250"of faceset is the right setting,am I wrong?

fix_modify AtC control thermal flux

fix_modify AtC initial temperature all 300
fix_modify AtC initial electron_temperature all 300

#add laser source
fix_modify AtC mesh create_nodeset bottom -INF INF -INF INF 0 0
fix_modify AtC fix temperature bottom 300
fix_modify AtC fix_flux electron_temperature bndy **

Q3:I read some previous reply about the units of this flux.to my understand,if power intensity is 10^16w/m^2,then 10^16 J/(sm^2)=10^16/(1.610^-19)ev/(sm^2)=(0.62510^35)/(10^1210^20)ev/(psA^2)=625ev/(psA^2),and in my simulation ,S=(53.62)^2,so the flux=625*(5*3.62)^2 ev/ps. if I am wrong about this point, Please correct me.

run 100
fix_modify AtC unfix_flux electron_temperature bndy

#boundary condition
#region boundary block 0 16.315 0 16.315 163.15 173.15 units box
#compute T internal temp/region boundary
#variable F equal -1.478e-3*sqrt(c_T)
#fix kick internal addforce 0 0 v_F
Q4:To get rid of the artifact from stress wave reflection in the target, a region of
10 Å along the z axis is built at the bottom of the target.But I don’t think it is necessary.

run 20000

1 Like

Maybe the ATC folks can advise on this (CCd)

Steve

Q1 yes
Q2 if the box you use to define the face set is coincident with any faces they will be included so if you want a portion of a plane all but one of the six numbers should be slightly off the coordinates of the mesh. You can check by issuing a
… Mesh output filename
Command and visualizing the results
Q3 atc uses energy units that are consistent with lammps mass time and length units ie typically different from lammps energy units
Q4 removing artificial reflections properly is quite involved. See jones and Kimmer prb 2010 for a method you can use with existing lammps commands.

Hope this helps

Reese Jones
rjones@…3…
925 294-4744

1 Like

Hi Emily, please keep the lammps user list cc’ed. Your input deck looks like it is applying the boundary flux only for 100 time steps, and I can’t tell just by eyeballing your number if the flux is in the correct units (see Reese’s Q3 response below). If you suspect that AtC is not acting per its documentation, what Reese is requesting is a small example problem that shows the discrepancy, e.g., if you think ATC is not adding the heat flux correctly please send us as small a problem as you can (few atoms, few finite elements) with as few extraneous commands in the input deck as possible. This will enable to us to examine the problem. However, I would also advise you double check your units and/or try applying the heat flux for a longer period of time.
Jeremy

hi,Jeremy
I think the flux unit is energy per area per time,e.s eV/A^2/ps,is that right? my laser pulse duration is 100fs,so I have to apply heat flux for 100fs. I also checked the output ,but I am not sure about it’s meaning.e.g"1 atomic_temperature_mean:2 atomic_temperature_std_dev:3 electron_temperature_mean:4 electron_temperature_std_dev:5 temperature_mean:6 temperature_std_dev"
the small example is as below:( attachment is potential and ttm file)
#测试小系统运行
#initialization
units metal
dimension 3
boundary p p s

lattice bcc 3.263
region MD block 0 5 0 5 10 20
region FE block 0 5 0 5 0 20
create_box 2 MD
create_atoms 2 region MD basis 1 1 basis 2 2

#set potential
pair_style eam/fs
pair_coeff * * CuZr_B2.eam.fs Cu Zr
neighbor 0.3 bin
neigh_modify every 10 delay 0 check yes
group internal region MD
timestep 0.001
thermo 200

velocity internal create 300 1234156
fix 1 all nvt temp 300 300 0.01
run 1000
unfix 1

#ttm-md coupling
fix AtC internal atc two_temperature CuZr_ttm.mat
fix_modify AtC mesh create 1 1 5 FE p p f
fix_modify AtC mesh create_faceset bndy box -inf inf -inf inf 10 20
fix_modify AtC control thermal flux faceset bndy

fix_modify AtC initial temperature all 300
fix_modify AtC initial electron_temperature all 300

#add laser source
fix_modify AtC fix_flux electron_temperature bndy 625
run 100
fix_modify AtC unfix_flux electron_temperature bndy
fix_modify AtC output atc_fe_output 100
run 10000

in my mind,the atom temperature will keep rising in first 10ps after adding the heat flux to electron_temperature field. (by the way , I use Cu’s data instead of CuZr in TTM for test)

sincerely
Emily

在 2014-08-20 05:59:21,“Templeton, Jeremy Alan” <jatempl@…3…> 写道:

CuZr_B2.eam.fs (2.16 MB)

CuZr_ttm.mat (301 Bytes)

Emily, as Reese pointed out in his earlier response, AtC units are different that LAMMPS units in that we use the primary variables, mass, distance, time, and charge, to get our derived units like energy (the one exception is electric potential, which is in units of electric field X distance). For metal units, energy, is mass * distance^2/time^2.
Jeremy

hi,Jeremy
Thanks to your professionalism and patience, I finally understood what is your meaning and solved the main problem. I still have some other questions:
Q1 I do not quite understand the output in AtC.e.s “1 atomic_temperature_mean:2 atomic_temperature_std_dev:3 electron_temperature_mean:4 electron_temperature_std_dev:5 temperature_mean:6 temperature_std_dev”
Q2 I want to get result like" spatial and temporal evolution of temperature and pressure". can the" mesh output" achieve my goal? or I should use fix ave/spatial(time) command?

Emily

在 2014-08-20 11:27:43,“Templeton, Jeremy Alan” <jatempl@…3…> 写道:

hi Emily

these are global scalars i.e. reductions of the corresponding field variables defined at every node. For instance

temperature_mean : the average temperature of the whole system, bar T =1/N sum_I T_I

temperature_std_dev : the standard deviation of the temperature

Reese

ps the timescale of the response of the phonon/atomic temperature to an electron heating source is strongly dependent on the electron-phonon coupling coefficient you use

Reese Jones
Sandia National Laboratories
P.O. Box 969, MS 9404
Livermore, CA 94551
(925) 294-4744 or 800 4SANDIA x2944744

hi, AtC folks
I met problem about the Material parameter file in TTM model. there are my questions:
1,in the file,there are 5 parameters: lattice heat capacity,lattice thermal conductivity,electronic specific heat,electronic thermal conductivity,electron_phonon couple coefficient. but I’m not sure about their units
2,what is the “constant” and "linear"really mean? in my mind,Ce=rTe,Ke=K0Te/Tl,and G is a constant value,why does it seem like the opposite in the file?

Emily

Constant means constant coefficient so the relation is linear. Likewise a linear coefficient gives a quadratic relation.

The units in the mat file are atc units ie mass length time consistent with lammps and energy self consistent

Reese Jones
rjones@…3…
925 294-4744

Emily, for the first question, can you attach your material file and specify which materials you have a question about? The answer to your second question is yes.
Jeremy

dear Jeremy
The material file I talk about is Cu_ttm.mat in the atc examples. because I want to simulate other material(Ni and CuZr),I have to figure out it’s true meaning. The contents of the file are as below:

material Cu metal
heat_capacity constant
capacity 0.211977459280654
end
heat_flux linear
conductivity 67.4479848
end
electron_heat_flux linear
conductivity 441.362750535
end
electron_heat_capacity constant
capacity 0.005817388689
end
electron_phonon_exchange linear
coefficient 0.0156575679
end
end

I try to change these datas’ units ,but they don’t match the data in literature. e.g Ce=r*Te, r=96J/m^3/k^2 if I change the units to lammps units ev/A^3/k^2, the result is not “0.005817……”(data in the file).
as to my second question last mail,you said “yes”. what is that mean?

Emily

Emily
the energy units for ATC are not eV. They are formed from the mass, length and time units
Reese

hi, I try to change these data to atc units(mass,length and time units), but they still don’t match at all. I‘m so
confused. can you provide me the source of these data (Cu_ttm.mat)? or some relevant literatures would help! thank you very much!

Please see the (Jones) paper listed on the fix atc page which will describe the theory as well as the parameters used for certain studies and their sources.
Jeremy