Browsing by Author "Van Tonder, John Dean"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    Using full field data to produce a single indentation test for fully characterising the mooney rivlin material model.
    (Stellenbosch : Stellenbosch University, 2024-02) Van Tonder, John Dean; Venter, Martin Philip; Gerhard Venter; Stellenbosch University. Faculty of Engineering. Dept. of Mechanical and Mechatronic Engineering.
    ENGLISH ABSTRACT: In the field of material characterization, a well-known problem in literature has been identified. The problem involves the solutions obtained from inverse Finite Element analysis when characterizing hyperelastic material model parameters, which are often non-unique. A gap in the literature exists regarding the handling of the non-uniqueness issue. Providing a solution for this has meaningful implications for engineering applications. The nature of these non-unique solutions is that they fit the dataset of the load case they were characterized on with indistinguishable errors from the actual optimal set of model parameters. These solutions prove to be sub-optimal when applied to load cases for which they were not characterized, failing to predict accurate material behaviour. The research presented in this dissertation addresses this non-uniqueness problem for the Mooney Rivlin model by introducing a novel contribution. The contribution involves a newly discovered concept known as hyperplanes, which manifest as flat, plane-like regions in the low-error regions of the design space. This discovery enables the isolating of a single, optimal set of material coefficients. The hyperplanes serve as the foundation for a new inverse Finite Element characterization method formulated as a constrained optimization problem. The main contribution of this formulation is that allows the the user to specify which loading state of the material deformation path they wish to fit. This is achieved by specifying specific measurement points. Additionally, this formulation allows for tolerances to be applied on these measurement points adding an additional level of compliance to the material characterisation. The behaviour of these hyperplanes was investigated, initially through a simulated indentation test that involved full-field digital image correlation experiments. However, these simulations provided a controlled environment to explore the characteristics of hyperplanes under noise-free conditions, leading to the development of the constrained optimization method. The applicability of the hyperplane concept was then validated using physical test data and compared with material testing standards. The results of this comparison study indicated that using hyperplanes in the inverse characterization process produced a more comprehensive set of material parameters than the test standards. In conclusion, this dissertation asserts the indispensable role of hyperplanes in isolating the true optimal set of Mooney Rivlin model parameters, thus addressing the identified gap in the literature and delivering a valuable contribution to the field of material characterization.