Example: sputtering of Zr(0001) target.
----------------------------------------

1734-atom Zr(0001) target bombarded by 5 keV Ar. 

Simulations terminate after 1000 fs, and output the coordinates
of emitted atoms (including the projectile).

The simulation model uses the WM2 potential from Table I of:

F. Willaime & C. Massobrio, Phys. Rev. B 43 (1991) 11653-11665.

[Note: the summation method used by WM differs from that used by
Kalypso, meaning that the WM values of A and q must be multiplied
by 2.0 in Kalypso.] This hcp potential has quite a long cut-off, 
compared with those normally used for fcc targets.

I have extended the cut-off from 5.4 A (WM value) to 5.5 A. This does not
affect the Zr lattice energy (the next shell is at 5.56 A), but it
gives a more realistic switching function force curve.

Kalypso reports a cohesive energy of 6.16967 eV at lattice site 723
(when vibrational effects are turned off in MDL file). WM fitted
the potential to the value 6.17 eV, but used a slightly different c/a
ratio for the lattice, which probably accounts for the discrepancy
(unless it is due to fitting error).

Inelastic & vibrational effects are included in this model.

The simulation model uses 105 impact points. The impact file is
created in a similar way to that required for a (111) hcp surface,
as described in the online Help.

The target size, and number of impact points have been kept at
modest levels in order to keep the project manageable.

The simulation yields a predicted sputter coefficient of 1.17. This
is estimated by filtering the output file, dynvars.snk, with
the expression [rw > 0] & [rz > 6E-10] (123 records satisfied this 
condition; 123/105 runs = 1.17 atoms per projectile). The 
experimental value is 1.12 [M.T. Robinson & A.L. Southern,
J. Appl. Phys. vol. 39 p.7 & p.3463 (1968)]. This agreement
is unusually good.

The prediction could be refined by excluding edge atoms, as 
described in the Simulation Primer, and by using a larger target.