2/2/2015 Authors: Ha-Rok Bae; Hiroaki Ando; Sangjeong Nam; Sangkyum Kim; Christopher Ha
J. Mech. Des.. 137(1), 011001 (2015); doi:10.1115/1.4027849

Typical data analytics obtain prediction models and data patterns by focusing on a numerical abstraction of data. In many cases, these results are a lack of physical interpretations and inappropriate to be directly used as design data for a new engineering design development. To address this technical gap, this study introduces an integrated framework of Engineering Data Analytics (EDA) in which the data transformation and featuring processes are coupled with physics-based model simulations. This paper demonstrates how EDA, integrated with damage-based pattern amplification, can help to identify the fundamental structural damage loads and optimize a sensor layout.  

For the Full Article see ASME’s Digital Collection