A simple, effective method is the ‘steepest descents’ method.

· Moves ‘straight downhill’

· Calculate the forces (gradients) on the atoms - the first derivative of the energy), first-order method

· The gradient is a vector (g), with one component for each coordinate. The step, s, is in the direction –g.

· The new coordinates, x_k+1, following the step from x_k are given by

x_k+1 = x_k + l_ks_k

l_k is the step size - chosen by iteration to find approximate minimum along the direction s.

Next: features of the steepest descents minimization method