Recent petrological, structural and geochronological studies of the eastern margin of the Bohemian Massif (Czech Republic) suggest a conceptual geodynamical model to explain exhumation of lower crustal (20 kbar, 800 degrees C) felsic rocks. The model involves indentation of a weak orogenic lower crust by an adjacent rigid mantle lithosphere, resulting in crustal-scale buckling of the weak orogenic lower/middle crust interface followed by extrusion of a ductile nappe over the rigid promontory. The hypothesis has been investigated using both analogue and numerical models. Analogue experiments using a three layer sand-silicone setup were carried out in Rennes laboratory (France). Results show that the most important features of the conceptual model can be reproduced: extrusion of lowermost silicone over the indenter and flow of horizontal viscous channel underneath a rigid lid above the actively progressing promontory. Furthermore, experimental results show that a plateau develops above the channelling lower crust. Two sets of sandbox-scale numerical simulations were performed. The first set of experiments is designed to study the influence of viscosity stratification within the crust on the extrusion process. A second set of experiments were performed in order to quantify the influence of the viscosity and the geometry of the indentor. Non-dimensional scaling laws were derived to predict the maximum extrusion rates associated with the indentation mechanism. Such laws enable the computation vertical extrusion rates that are in good agreement with natural exhumation rates inferred from petrological data. Finally, we discuss the potential positive feedback of Rayleigh Taylor instability on vertical extrusion for the case of Eastern Bohemian Massif