Damp coefficient for different types of beads in a polymer chain

Dear all,

I am trying to simulate a polymer with two types of beads and each of different masses and sizes. To calculate the damp coefficient, I used the formula from the documentation,

damp=mass/gamma; where gamma=3 x pi x n x d.

The value of damp for two different particles is 2 x 10^5 for a particle with a mass of 179000 Da and 100 Angstrom size and 1.5 x 10^3 for a particle with a mass of 6500 Da and 20 Angstrom size.

But the documentation describes the particle with a small damp coefficient as more viscous. Can anyone please suggest me what damp to be used for the particles of the above size?

Hi @RakeshK,

I think you are confusing both documentation of the fix viscous and fix langevin commands on the one hand and the physics on the other hand. This is not directly related to LAMMPS and more concerning the general way to setup the model for your simulation.

Some comments I can provide to help:

• The formulas you are using are physics formula. They are external to LAMMPS and are the basis on which the commands are programmed. You should use the command to make a model you define beforehand, not the other way around.
• The formula you are using, \gamma=3\pi\eta d, is valid in the context of Brownian dynamics as stated in the documentation. The value of \gamma depends on an external parameter that is the viscosity of the surrounding fluid. This is equivalent to the expected diffusion coefficient of the Brownian particle (since they are invert proportional to one another). Brownian dynamics and Langevin dynamics are also two different dynamic models. Therefore, what is the meaning of your calculation?
• The size of the particles is only taken into account with finite-size particle models (sphere and ellipsoids). Are you planning on using some? It is totally possible to model polymers using bonded point particles. You should refer to the literature and define what best suits your aims. Then use the setup and parameters are provided.

You might get on track easier through more reading of the scientific literature related to polymer modelling using MD and advises from skilled colleagues or advisor around you. This might help you define your problem and its solutions better.

1 Like

hi @Germain,

So, If simulate a bead-spring polymer with two different types of beads (A and B); sigma_A=10nm and sigma_B=2nm.

If I am simulating them using fix langevin, how do I calculate its damp coefficient parameter,
while the viscosity of the solvent is known?

Do I have to provide different damp values for the different types of particles?

If youâ€™re using the fix langevin command the straight forward way to apply different damp factor is the use of the scale keyword. As stated in the manual:

The keyword scale allows the damp factor to be scaled up or down by the specified factor for atoms of that type. This can be useful when different atom types have different sizes or masses. It can be used multiple times to adjust damp for several atom types. Note that specifying a ratio of 2 increases the relaxation time which is equivalent to the solventâ€™s viscosity acting on particles with \frac{1}{2} the diameter. This is the opposite effect of scale factors used by the fix viscous command, since the damp factor in fix langevin is inversely related to the factor in fix viscous. Also note that the damping factor in fix langevin includes the particle mass in Ff, unlike fix viscous. Thus the mass and size of different atom types should be accounted for in the choice of ratio values.

So for example if the damp factor acting on type 2 is half as much as the one acting on type 1 you would have a command like:

fix all LAN langevin ${Temp}${Temp} ${damp}${random} scale 2 0.5


assuming you define variables with the desired temperature values and damping parameters in ${Temp} and ${damp}. As for the value themselves this is a question that depends on the model you are using and the properties you want to reproduce. This is something that you have to carefully choose based on the literature and the physics you want to simulate. I do not know what your model is and what youâ€™re trying to simulate so I cannot provide any insight more useful than that.

2 Likes

Hi,
I will explain my system. I am coarse-graining a DNA as a bead spring polymer with two types of beads, namely nucleosome bead or type A (sigma=100Angstrom) with a mass of 180000 Da and linker DNA bead or type B bead (sigma=20A) with a mass of 6500 Da.

Well, fix langevin imposes a frictional force of F = - mass * v / (damping time); Stokesâ€™s law says the frictional force on a spherical particle in laminar flow is 6 \pi R \eta v; if you equate the two, you should be able to determine the appropriate damping time, in the appropriate units.