Various industrial processes involve materials consisting of grains. The spectrum ranges from coarse grained matter and fine powders to nano-powders. As grain size decreases, the influence of cohesion increases leading to an enormous pore volume, which sometimes can exceed 90%. One main goal of this thesis is to understand this macroscopic behavior on a microscopic level and to quantify it. By using the contact dynamics method we simulate the evolution of an ensemble of grains in two dimensions. An extension dealing with cohesion and rolling friction turns out to be crucial for explaning the pore stabilization mechanism. Furthermore, our model shows that strong cohesion can stabilize even low coordinated packings such as particle chains. After compaction the porosity depends on the ratio of applied pressure to the cohesive force (scaled by the particle radius). For high pressures this dependence can be described by a power law. Consolidation experiments by the group of Prof. Schwedes at the Technical University of Braunschweig confirm this behavior for different sized powders. The influence of this scaling function on the stress driven compaction process can be explained by a simple model. Extending this model to strain driven processes, we predict density inhomogeneities inside the powders, which diminish with decreasing deforming rate and vanish in the limit of a quasistatic process. Experiments, usually performed at low strain rates, validate this homogeneous deformation in this case. Furthermore, we investigate the influence of the initial configuration. Fractal structures which include high differences in local density turn out to better stabilize pores against applied pressure. This is in accordance with experiments on silica gels which show increasing mechanical strength with growing local density inhomogeneities.