I have derived a number of many-body potentials using the Finnis-Sinclair approach which is a second-moment approximation to the tight binding approach in which only near-neighbour hops are considered - equivalent to representing the electronic band by only its width.

The method is superior to pair-potentials, and similar in accuracy and complexity to the Embedded Atom Method and the effective medium approach.

Implementation of these potentials is the same as for the embedded atom method.

Potential parameters for aluminium, copper, gold, iron, nickel, palladium, silver, titanium and zirconium can be obtained from G.J.Ackland@ed.ac.uk. They are all well tested and I can give you some indication of the work already done with them.

Alloy parameters for the Al-Ti, Cu-Ti, Ni-Al, Au-Cu, Au-Ag, Cu-Ag Cu-Fe FeP and FeC systems also exist. Mail me for details.

If the system you want is not listed above, try asking me nicely and I'll see what I can do.