search

radical-cybertools / radical.repex.at

This is the github location for RepEx developed by the RADICAL team in conjunction with the York Lab.
Other
4 stars 3 forks source link

T dimension issue in TUU #36

Closed haoyuanchen closed 8 years ago

haoyuanchen commented 9 years ago

As a follow-up of the issue I brought up in today's meeting (7/27), there're exchanges happening in T dimension if replicas differ in temperature by only 1 K. However, when I increase the gap to 5 K, no exchanges in T dimension happened, which is against the physical intuition.

Then I found that the energy differences (from the matrix column files) for T dimension is orders of magnitude larger than for U dimension, which resulted in the fact that virtually no exchanges in T dimension could be accepted.

There could be something wrong in the temperature exchange codes. I'll continue to investigate this.

haoyuanchen commented 9 years ago

Just an update, the T dimension exchange seems to work normally when I run the ace_ala_nme example. The previous issue was observed when running the phos_trans example. So it might depends on the real physical system. I'll continue to investigate this and, if you observed different exchange behaviors, please let me know.

antonst commented 9 years ago

Thanks for the update! I will look into this as well.

haoyuanchen commented 8 years ago

With the current code, the temperature exchange seems to be fine.