add more options to cell_dofree to fix cell paremeters and some fixes in vcsubs for cell_dofree.
I added the following options to cell_dofree: fixa, fixb, fixc, a, b, and c. fixa fixes the first cell vector and relaxes 2 and 3. Similarly for fixb amd fixc. a fixes the avec(1,1). b fixes avec(2,2) c fixes(3,3) . Now actualy what is done is to zero the corresponding stresses. Now for some reason I don't understand, using bfgs there are still some small undesired changes in parameters that should be fixed. However, with my fixes with damped dynamics they are fixed correctly. The convergence test for damp-w though with fixed parameters is unsatisfactory--there will be always be non-zero stresses since parameters are fixed. Ronvcrelax.in