Damping vibrating cantilever beam

Hi, I have the following question:

Say, I applied an initial velocity on the free end of a cantilever beam, and then released the beam. I also fixed a temperature of 300 K. The beam keeps vibrating for a very long time. What I want to do is to damp the beam after some time in order to analyze deformation (if there is any). How can I successfully damp the beam so that it will come into rest about its initial position? I need the system still have 300 K temperature.

Thanks,

Hi, I have the following question:

Say, I applied an initial velocity on the free end of a cantilever beam, and
then released the beam. I also fixed a temperature of 300 K. The beam keeps

300K including the motion of the cantilevel itself or without?

vibrating for a very long time. What I want to do is to damp the beam after
some time in order to analyze deformation (if there is any). How can I

you can always dissipate kinetic energy using fix viscous or use fix
addforce with a atom style variable to apply a position and velocity
depending force counteracting the motion.

successfully damp the beam so that it will come into rest about its initial
position? I need the system still have 300 K temperature.

why would you need to thermalize it? what would be thermalizing the
object in a real experiment?

axel.

Hi, I have the following question:

Say, I applied an initial velocity on the free end of a cantilever beam, and
then released the beam. I also fixed a temperature of 300 K. The beam keeps

300K including the motion of the cantilevel itself or without?

vibrating for a very long time. What I want to do is to damp the beam after
some time in order to analyze deformation (if there is any). How can I

you can always dissipate kinetic energy using fix viscous or use fix
addforce with a atom style variable to apply a position and velocity
depending force counteracting the motion.

successfully damp the beam so that it will come into rest about its initial
position? I need the system still have 300 K temperature.

why would you need to thermalize it? what would be thermalizing the
object in a real experiment?

axel.

Sounds like you are an engineer trying to do physics. Is your beam atomistically described?
Please advance a bit more details about the nature of your system.
Carlos

Hi, I have the following question:

Say, I applied an initial velocity on the free end of a cantilever beam, and
then released the beam. I also fixed a temperature of 300 K. The beam keeps

300K including the motion of the cantilevel itself or without?

vibrating for a very long time. What I want to do is to damp the beam after
some time in order to analyze deformation (if there is any). How can I

you can always dissipate kinetic energy using fix viscous or use fix
addforce with a atom style variable to apply a position and velocity
depending force counteracting the motion.

successfully damp the beam so that it will come into rest about its initial
position? I need the system still have 300 K temperature.

why would you need to thermalize it? what would be thermalizing the
object in reality?

axel.

Thank you for the responses.
Axel: I am trying to simulate a real experiment at room temperature (at a much smaller scale). That is why I need 300 K. I am using temp/berendsen to fix the temperature, so I guess it includes the beam vibration. I can try fix/viscous, but I suspect that it will slow down the system at an unrealistic state. In this respect, I wonder if the most realistic case is to have a fix/viscous along with a gravitational force? At the end, what it seems realistic to me is to have the cantilever beam come into rest about its initial position.
Carlos: This is an atomistic system consisting of 100K atoms in a cantilever beam configuration where the left end is fixed at all dimensions.

Well, if you are trying to do atomistic MD at finite temp then you should know that there is no way you can set the atoms at rest and expect to mimic temp effects. Besides, you’ll get nothing out of using gravitational forces (at least without boosting like crazy the strength of the field) to drive the system. those forces are tiny in any real life experiment. what you call “rest” if just the average motion of the system. You could have temp and rest at the same time if you learn how to withdraw what you call displacement in real life from what atomic motion is at the micro level. Which approach is best (no temp + action + damping vs temp + action) depends probably on your system size and the physics you are trying to capture.

Carlos

Thank you for the responses.
Axel: I am trying to simulate a real experiment at room temperature (at a
much smaller scale). That is why I need 300 K. I am using temp/berendsen to
fix the temperature, so I guess it includes the beam vibration. I can try
fix/viscous, but I suspect that it will slow down the system at an
unrealistic state. In this respect, I wonder if the most realistic case is
to have a fix/viscous along with a gravitational force? At the end, what it
seems realistic to me is to have the cantilever beam come into rest about
its initial position.

i disagree. while you do a real experiment at a room temperature
environment. there is *nothing* that will couple the entire system to
an external heat batch like the various thermostat options do. those
are meant to model thermalizing for a system *immersed* in a large
bulk. that is most certainly not the case for your system. even if you
do not perform a experiment in vacuo, the interactions with gas
molecules would be so rare, that they can be neglected (unless you
have a a rather high pressure.

thus the reasonable approach to model your system would be to
equilibrate it while in rest to the desired temperature (and better
use a proper thermalization method (i'd recommend fix langevin) and
not something that is only acceptable for bulk liquid systems like
temp/berendsen. then you remove the thermalization, integrate with fix
nve and add the "kick" that makes your cantilevel vibrate.

please sit down, take a piece of paper and compute the gravitational
force on your system and compare it to typical forces that your atoms
experience. you should see that it is far below the error margin of
any typical forces and thus useless to consider for a system of your
size.

finally, if you want to artificially slow down your system to come to
rest, i wonder what is left as a meaningful result. how would the
system dissipate its kinetic energy in reality? how can you study
damages, by artificially shortening the time your system is subject to
damage? wouldn't it be more meaningful to just let it go until you see
it come to a rest by itself? or just look at it for a given time as it
is and then observe the deformation (how?) and extrapolate to a longer
period?

axel.

p.s.: sorry for the multiple replies that seemed to have gone out,
even through gmail told me that mails were not sent due to a server
error.

please sit down, take a piece of paper and compute the gravitational
force on your system and compare it to typical forces that your atoms
experience. you should see that it is far below the error margin of
any typical forces and thus useless to consider for a system of your
size.

Wouldn't that be nice? (the strong gravitational pull I mean) You just had
to pick your favorite model, go say hi to her and then BAM!!!.. Until death
pulls you apart :wink:

Carlos

I am not expecting atoms at rest and temperature at the same time as it does not make sense as you pointed out. I am sorry I was not clear on that. I am trying to get the beam at rest such that it stops vibrating up and down (a rest at a larger scale). And, I understand that gravitational forces are not comparable to the forces I have in the simulation, so that was a bad idea :slight_smile: I think the connection from atomistic motion to macroscopic level in my case is related to deformation type: elastic or plastic. Elastic deformation is the recoverable deformation while plastic is not because of some reconfiguration of atoms. As such, after I apply the initial velocity, I want to see if that caused a plastic deformation. For that, I need the system to recover elastic deformation and left with plastic deformation, which I thought is possible if the system comes into a rest (at the larger scale).

I am not expecting atoms at rest and temperature at the same time as it does
not make sense as you pointed out. I am sorry I was not clear on that. I am
trying to get the beam at rest such that it stops vibrating up and down (a
rest at a larger scale). And, I understand that gravitational forces are not
comparable to the forces I have in the simulation, so that was a bad idea :slight_smile:
I think the connection from atomistic motion to macroscopic level in my case
is related to deformation type: elastic or plastic. Elastic deformation is
the recoverable deformation while plastic is not because of some
reconfiguration of atoms. As such, after I apply the initial velocity, I
want to see if that caused a plastic deformation. For that, I need the
system to recover elastic deformation and left with plastic deformation,
which I thought is possible if the system comes into a rest (at the larger
scale).

wouldn't it be better in that case to just look at an arbitrary chunk
of your material and simply determine when you have the transition
from elastic to plastic deformation? you would have much better
control of your simulation *and* have it much easier to extract useful
information. just forcing your system to arbitrarily do what you want
it to behave like through external manipulation is more like making a
movie and less like doing a scientific experiment.

axel.