Re: Huge Energy Discrepancy (Order of Magnitude) Between Zero-Field Cooled State and Subsequent Warming Simulation

30 views

Skip to first unread message

Message has been deleted

Richard Evans

unread,

Aug 26, 2025, 2:25:51 AM (12 days ago) Aug 26

to Vampire Users

Dear Christopher,

This might be due to average vs instantaneous energy (positive indicates unstable, but maybe that's a typo?). Do you calculate the instantaneous (output:total-energy) as well as average (output:mean-total-energy)?

Cheers,

Richard

On Tuesday, August 26, 2025 at 5:46:36 AM UTC+1 tirumis...@gmail.com wrote:

Dear VAMPIRE Development Team,
I am currently running a simulation on a 2D magnetic material, involving a zero-field cooling process followed by warming, and I've encountered a puzzling phenomenon while analyzing the energy results. I would appreciate some guidance from the experts here.

My simulation procedure is as follows:

Step 1: Zero-Field Cooling (ZFC)

I started from a random spin state at a high temperature of 600K and slowly cooled the system down to 0K under zero magnetic field. The objective was to obtain the ground-state magnetic structure.

After the simulation, I saved the final magnetic configuration at 0K and recorded its corresponding total energy, which I'll call E_0K (-6.54914e-16).

(File details: Input was input1_cooling, output was output_cooling, and the magnetic structure was spin_config_00000319_cooling.)

Step 2: Warming Simulation

Next, I loaded the 0K magnetic configuration obtained in the first step as the initial state for a new simulation. The goal was to plot the magnetization vs. temperature curve and observe the evolution of the magnetic structure with temperature.

In this warming simulation, the system was equilibrated for 2 million steps at each temperature point to ensure it reached thermal equilibrium. Following this, I sampled and averaged the system's energy over the next 200,000 steps.

After completing the calculation at the 0K temperature step in this warming run, the average energy was E_T (1.1496e-15).

The Problem:

These two energies, both calculated for the 0K state, differ by an order of magnitude. What could be the reason for this significant discrepancy?

Thank you very much for taking the time to read my question. Any explanations of the physical mechanism, suggestions for simulation methods, or diagnostic tips would be extremely helpful!

Thank you all!
Best regards, Christopher Johnson