This project aims at predicting and choosing between several alternative crystal structure models by optimizing with plane wave wave DFT methods. Calculating externally measurable physical properties may help to distinguish otherwise equivalent crystal structure models.Improvements of low quality structures from powder diffraction data is one of the main goals for the present project. Several obstacles of structure determination is for example the inability to determine hydrogen positions, whether it is carbohydrate structures or metal hydrides. With DFT calculations one should be able to test several possible structures and configurations regarding hydrogen atom positions. An altogether new structure may also be checked for plausibility in order to invest the time for chemical synthesis and laboratory exercises with real chemicals, i.e. predicting possible compounds. Another possible obstace would be to check several possible crystal structures before attempting to analyse diffraction data nd other type of measurements.