About compute ti

Hi,
I try to use compute ti to calculate the free energy.
I used a small system to check this command. I put some particles into a box which have a wall across the box.
The box is cut into two regimes. Particles are stay in one regime. And then i remove the wall, they go around inside the box.
I want to get the free energy change after the wall is removed. Theoretical approximation of the free energy change is : F(no wall)=ln2 F(wall).
But i found the free energy is totally the same. The main input file is posted here:

pair_coeff * * 0.5 3.0 3.36
variable flambda equal 0.1
variable dflambda equal 1
fix 4 all adapt 1 pair lj/cut epsilon * * v_flambda scale no reset yes
compute 1 all ti lj/cut v_flambda v_dflambda
thermo_style custom step temp press etotal pe ke c_1

I use f(lambda)=lambda . So the lambda value is changed from 0 to 1.

So what is wrong with my calculation?? Any help is appreciated! Thanks!

Suzan

I'll let Sai comment on this.

Steve

The free energy change you are looking for cannot be computed using compute/ti+fix/adapt. You are trying to compute the free energy change associated with pressure-volume changes in the system. This can be computed as Delta G = int VdP. Compute/ti with fix/adapt will allow you to compute free energy differences between states which differ in their energetic interactions, and any entropic differences arising out of such energetic interactions.

Sai.