Structure formation is a main soil forming process. It is a result of the reorganisation of particles. The understanding of the processes, which lead to the separation of particles due to shrinkage and hence cracking is based on the understanding of the controlling parameters for volume change. The rearrangement of particles in a porous material like soil is initiated by volume change caused by either external or by internal forces. Internal stresses may be considered as self-organising stresses, where particles are re-arranged for thermodynamic reasons. Internal stresses or tensile stresses are a result of changing water contents or water potentials, respectively. Generally, these changes cause volume change due to swelling and shrinkage. While volume change as a result of a change in the water status is usually related to changes in water content and described by the shrinkage curve, little is undertaken to define shrinkage by its dependent stress state variable water potential. It has been shown previously that it is possible to use empiric continuous functions, which are based on water potential as its variable, to model volume change. When combining this relationship with the water retention curve, the shrinkage curve based on the variable water potential can be calculated. Using continuous functions to predict the shrinkage curve allows the interpretation of these functions and the derivation of characteristic values defining and describing the mechanical and hydraulic behavior of soils, using water potential as the main variable. In the presentation it will be shown, that the shrinkage curve can be underlain by the water potential, i.e. by creating a volume-water content-water potential surface. Several mechanically related properties can be derived, which allow a parametric characterisation of the shrinkage curve and give an indication for the structural strength of a soil substrate. Furthermore, it allows to model the dynamic volume change of soils (shrinkage and swelling), which can be of great relevance for the modelling of the hydrology of porous media.