An Investigation into Deformation and Failure in Thin Sheet Electrical Steels Using a Non-Local Modelling Approach

Electrical machines are typically manufactured using stacked thin laminations of ferromagnetic material as a core in order to guide and amplify magnetic flux. Silicon steel, also known as electrical steel, is one of the most commercially successful materials for this role. This is due to its high magnetic permeability, low coercivity, and high resistivity, as a result of the alloying elements and large grained microstructure. The required geometries for rotors and stators are manufactured using the blanking process, which introduces edge damage and degrades the magnetic properties causing increased energy loss. Understanding the effects of the various parameters on edge damage is vitally important to guide and improve the blanking process. In this work, a novel non-local damage model is developed to predict edge damage resulting from this process. The model is an extension of the Bao-Wierzbicki type fracture model, which performs well due to its sensitivity to stress state. The difficulties in calibrating a damage model, such as the high variance of the fracture dynamics due to the size effect, have been addressed. It was found that calibrating the model based only on the limited geometries manufacture from sheet stock was possible through highly localised strain measurements from digital image correlation (DIC). To investigate low triaxiality samples, a device compatible with DIC for preventing out of plane deformation due to buckling is demonstrated. A new method for in-situ damage detection is presented using hall effect sensors to detect the change in magnetic flux flowing through a sample. The model is validated using a number of geometries and load conditions and found to reproduce fracture mechanics accurately.