A designer of a new mechanical ice protection system for airplanes needs to know how much and in which way he has to deform the surface to break off the ice. The ice adhesion strength is often used as a design value. Several methods have been published to measure the adhesive strength of ice. This paper analyzes the interface stresses created by those methods and discusses the way the adhesion strength is derived. A finite element method tool is used to provide insight into the stress state for different load cases. The implication of these illustrations is that equations which use only ultimate force and total interfacial area to calculate adhesion strength miss local stress concentrations and crack nucleation. Hence, the derived adhesion strength may not be comparable within different testing methods, because each testing procedure neglects different parameters like specimen size, substrate thickness and stiffness. In addition the ice adhesion strength is different if peel stress or homogeneous shear stress is applied.Adhesive joint theory can be used to show how to model stress concentrations. Mentioned theory explains why thick ice layers break off more easily, and bending is the most efficient load case to remove ice. Preliminary concepts for new designs of mechanical IPS are given by improving the understanding of stress concentration.