colvars: force constant for harmonic

Dear users,

I have set up a simulation for estimating the free energy using Umbrella sampling.

I am in need off following suggestions.

  1. When I specify the width=0.05 and force constant=50.

The force constant for colvar “xxx” will be rescaled to 20000 according to the specified width.

If I delete the width line, force constant is not being scaled. Is the free energy obtained from this simulation ( without width) is reliable one??

  1. How can be use different force constant values for different stages in US?

At minimum FE, K is small & at maximum FE, K being high?

Thanks in advance,


Hello Spandana, the scaling feature exists to let you control multiple variables with different dimensions. You are free to use a width of 1 if you like.

It sounds like you are planning on using umbrella sampling without having a clear idea of how the restraint force constant correlates to the width of the sampled distribution.

My recommendation, aside from reading the literature, is to try a pmf calculation for a very simple example, such as two Lennard Jones atoms interacting in a box of fixed size.