Question on compute temp and compute temp/deform

Hi All,

I have a question about biases in thermostats due to streaming velocities of atoms.

If I’m deforming and thermostatting the box as follows:

“fix fxnvt all nvt/sllod temp 298 298 100.0
fix fxdfm all deform 1 xy wiggle {ampl} {tperiod} units box remap v”

and then calculate the temperature:

“compute cmp0 all temp”

Will the temperature calculated be the same as that calculated by fix nvt/sllod?

I ask this because a section in the manual states that incorrect use of compute temp/deform actually adds the bias back in to the computation, rather that takes it out.

Regards,

Tim.

Hi All,

I have a question about biases in thermostats due to streaming velocities
of atoms.

If I'm deforming and thermostatting the box as follows:

"fix fxnvt all nvt/sllod temp 298 298 100.0
fix fxdfm all deform 1 xy wiggle \{ampl\} {tperiod} units box
remap v"

and then calculate the temperature:

"compute cmp0 all temp"

Will the temperature calculated be the same as that calculated by fix
nvt/sllod?

​no.

axel.​

This is explained on the fix nvt/sllod doc page (likewise for other

thermostatting fixes). It creates its own compute temp/deform.

You can query the output of that compute (or multiple compute temp
comands) in your thermo output.

Steve