This thesis presents a systematic analysis of locking phenomena in the context of finite elements and methods for their elimination. The Discrete Strain Gap method (DSG) represents a universal concept for the elimination of geometric locking effects. Exceptional characteristic of this formulation is the uniform treatment of all kinds of geometric locking effects. It is applicable to elements of arbitrary shape and order without the determination of collocation points. Comparison with established element formulations as well as numerical examples confirm the efficiency of the proposed concept. Analysis of locking phenomena and the understanding of their origins are essential for the assessment of finite element properties. Beside the definition of the term locking, mechanical, numerical and mathematical interpretations of locking effects are presented. Special attention is devoted to a strict separation and definition of the different locking phenomena. Furthermore, this thesis deals with the application of the Discrete Strain Gap method (DSG) and its suitability concerning the described locking phenomena. The proposed DSG modifications are applicable for both three-dimensional solid elements and structural elements. Description of well-known concepts like Reduced Integration, Assumed Natural Strains (ANS) and Enhanced Assumed Strains (EAS) provides a framework suited for a comparison to the modifications introduced by the DSG method. Starting from the Hu-Washizu principle the development of a variational basis leads to a mathematical foundation of the DSG method. The DSG method represents a uniform concept for the formulation of finite elements. Based on a variational principle, this method convinces by its excellent numerical performance and in particular by its universality and conceptional clarity.
«