Development of a method for the efficient and contactless material parameter identification of string instruments using parametric model order reduced finite element models
Final Report Abstract
In this research project, a method was developed for the efficient identification of material parameters in string instruments. The foundation for this was an extremely detailed finite element model, which enabled the identification of a guitar’s material parameters through a model updating procedure. Typically, such methods are computationally intensive and require a very large number of model evaluations. The resulting high computation time was reduced to approximately one-thousandth by using efficient surrogate models, which were created through a combined approach of projection-based parametric model order reduction and purely data-driven methods. The method was successfully applied to distinguish between two guitars based on their material parameters. Beyond the knowledge of the pure material parameters themselves, the resulting models are also of particular significance. For instance, the finite element model of a complete guitar showed very good agreement with measurements, and in the case of a guitar soundboard, it was demonstrated that the models with identified material parameters are capable of predicting the effect of geometrical modifications. With the completion of this research project, a methodology is now available that enables the reliable identification of material parameters using nondestructive measurements on guitars in combination with a finite element model of the instrument. This method also allows for the identification of differences between guitars that are assumed to be identical based solely on their geometry. The project consisted of three essential components: 1) The complete development of the measurement methodology and data evaluation, 2) The creation of a finite element model including the assessment of key influencing factors, 3) The development of an optimization process to optimally adapt the material parameters of the numerical model to measurement data, thus enabling their identification. In addition to the planned objectives of the approved project proposal, a feasibility study was conducted on a braced guitar top, demonstrating that certain natural material variations can be compensated for by modifying the geometry of the bracing. This opens up the visionary possibility of producing acoustic copies of instruments despite material variability. During the course of the project, seven publications were produced in which the project team presented its findings and ideas to the public, as well as one completed doctoral dissertation.
Link to the final report
https://doi.org/10.34657/27625
Publications
-
Numerical Models for Classical Guitars with Updated Parameters From Experimental Data. In Proceedings of Forum Acusticum 2020, Lyon
Brauchler, A.; Ziegler, P. & Eberhard, P.
-
An entirely reverse-engineered finite element model of a classical guitar in comparison with experimental data. The Journal of the Acoustical Society of America, 149(6), 4450-4462.
Brauchler, Alexander; Ziegler, Pascal & Eberhard, Peter
-
Modeling the Body of a Classical Guitar Using the Finite Element Method and Experimental Modal Analysis. In Fortschritte der Akustik - DAGA, Vienna
Brauchler, A.; Ziegler, P. & Eberhard, P.
-
A data-based method enhancing a parametrically model order reduced finite element model of a classical guitar. Proceedings of the 10th Convention of the European Acoustics Association Forum Acusticum 2023, 6241-6248. European Acoustics Association.
Cillo, P.; Brauchler, A.; Gonzalez, S.; Ziegler, P.; Antonacci, F.; Sarti, A. & Eberhard, P.
-
Distinguishing geometrically identical instruments: Possibilistic identification of material parameters in a parametrically model order reduced finite element model of a classical guitar. Journal of Sound and Vibration, 535, 117071.
Brauchler, Alexander; Hose, Dominik; Ziegler, Pascal; Hanss, Michael & Eberhard, Peter
-
Material Parameter Identification of Classical Guitars by Means of Surrogate Models Based on Parametric Model Order Reduction. In Fortschritte der Akustik - DAGA, Stuttgart
Brauchler, A.; Ziegler, P. & Eberhard, P.
-
Model-predicted geometry variations to compensate material variability in the design of classical guitars. Scientific Reports, 13(1).
Brauchler, Alexander; Gonzalez, Sebastian; Vierneisel, Manuel; Ziegler, Pascal; Antonacci, Fabio; Sarti, Augusto & Eberhard, Peter
-
Predictive Computational Models of Classical Guitars: Modeling, Order- Reduction, Simulation and Experimentation. Dissertation, Schriften aus dem Institut für Technische und Numerische Mechanik der Universität Stuttgart, Vol. 78. Aachen: Shaker Verlag, 2023
Brauchler, A.
