In any design, the dynamic characteristics of a part are dependent on its geometric and material properties. Identifying vibrational mode shapes within an iterative design process becomes difficult and time consuming due to frequently changing part definition. Although research has been done to improve the process, visual inspection of analysis results is still the current means of identifying each vibrational mode determined by a modal analysis. This research investigates the automation of the mode shape identification process through the use of parametric geometry and machine learning.In the developed method, displacement results from finite element modal analysis are used to create parametric geometry which allows the matching of mode shapes without regards to changing part geometry or mesh coarseness. By automating the mode shape identification process with the use of parametric geometry and machine learning, the designer can gain a more complete view of the part's dynamic properties. It also allows for increased time savings over the current standard of visual inspection
College and Department
Ira A. Fulton College of Engineering and Technology; Mechanical Engineering
BYU ScholarsArchive Citation
Porter, Robert Mceuen, "Application of Machine Learning and Parametric NURBS Geometry to Mode Shape Identification" (2013). All Theses and Dissertations. 5744.
NURBS, machine learning, parametric geometry, mode shape identification, automation