Before starting the simulation, let's have a look inside the script first. It's plain text, so open in your favourite text editor. You should see something like this:

<param>
   <restart = 0>
   <dt = 1.00e-14>
   <kT = 4.11e-21>
   <check = 10000>
   <num_steps = 1.000000e+11>
   <rng_seed = time>
   <trajectory_out_fname = emd_5043_8ang.ftj>
   <measurement_out_fname = emd_5043_8ang.fm>
   <vdw_forcefield_params = emd_5043_8ang.lj>
   <epsilon = 1.00e-02>
   <max_iterations_cg = 1000>
   <kappa = 1.00e+09>
   <epsilon_0 = 1.00e+00>
   <dielec_ext = 1.00e+00>
   <calc_stokes = 1>
   <stokes_visc = 1.00e-03>
   <calc_vdw = 1>
   <vdw_type = steric>
   <vdw_steric_factor = 0.010000>
   <calc_springs = 0>
   <calc_noise = 1>
   <calc_es = 0>
   <es_update = 1>
   <es_N_x = -1>
   <es_N_y = -1>
   <es_N_z = -1>
   <move_into_box = 1>
   <sticky_wall_xz = 0>
   <wall_x_1 = PBC>
   <wall_x_2 = PBC>
   <wall_y_1 = PBC>
   <wall_y_2 = PBC>
   <wall_z_1 = PBC>
   <wall_z_2 = PBC>
   <es_h = 3>
   <num_blobs = 1>
</param>

<system>
   <blob>
      <motion_state = DYNAMIC>
      <topology = emd_5043_8ang.top>
      <material = emd_5043_8ang.mat>
      <stokes = emd_5043_8ang.stokes>
      <pin = emd_5043_8ang.pin>
      <nodes = emd_5043_8ang.node>
      <surface = emd_5043_8ang.surf>
      <vdw = emd_5043_8ang.vdw>
      <solver = CG_nomass>
      <scale = 1.00e-10>
   </blob>
</system>

The FFEA input file is a structured file that has all the information that FFEA needs to run a simulation. Full detail on the algorithms and options available can be found in the FFEA runner section, including the input file reference, which will also allow setting up more sophisticated simulations. For demonstrating purposes, we need to change a few global parameters. So, for now, go ahead and set dt = 1e-15, num_steps = 1000 and check = 10. This will set us up for a run of 1 nanosecond, output a frame every 10 simulation time steps which are 0.01 nanoseconds in length.

Before starting the FFEA runner, you may want to consider setting your OMP_NUM_THREADS environment variable. The recommended number of threads is between 4 and 16. 16 will produce the best results, but will only be around 30-40% faster than running on 4 threads. For now, we will set

export OMP_NUM_THREADS=4

Now, let's run ffea!

ffea emd_5043_8ang.ffea

Once the simulation has finished (which should take less than 5 minutes), we can have a look at the structure in the viewer, and also perform some analysis. A small note; if a simulation crashes with an error talking about elements inverting, not to worry! The simulation time step has been set just a little high. Set dt to a smaller value, and rerun.

Using ffea as FFEA_runner was the most appropriate binary this time because we were simulating a single blob. Alternatively ffea_mb would perform better when using many blobs.