This thesis and the research work surrounding it, is oriented towards finding a solution to a problem in obtaining accurate material constants whenever only a single data set (i.e. uniaxial tension test data) is available in hyperelestic material characterization.
To begin with, the serious nature of the problem was highlighted through results of set of experiments. There, several material models were tried with two data fitting methods and the inaccuracy of data fitting with single data set could be proved beyond doubt from this exercise.
At the next stage, as a preliminary solution to the problem, a suggestion was given in the way of secondary data set generation from available data. The question at this point was about the method which could be adopted to generate the second data set. As an initial trial, exponential function was used with several exponents in order to generate data which could be consequently used as biaxial data. Amid some minor discrepancies, method delivered some promising results.
Second approach was sorted in order to get a better trajectory for the generated biaxial data. In this method, initial uniaxial data set was divided in to two segments and each segment was differently addressed. As a result, the trajectory of generated data nearly resembled the real biaxial data. Data fitting preceded the data generation, provided very encouraging results too. However, method had some serious shortcomings such as, unit incompatibility, and lack of use of uniaxial data in the later half. Due to these reasons, the method was not further examined for the use in the work.
Final experiments were done with six materials. Base material and other constituents were different in each of these cases and it resulted in varied data distributions in both uniaxial and biaxial data. An exponential function was once again used with a different exponent in finial tests. Proximity of generated data against real biaxial data was statistically tested. For the testing, 95% confidence interval was selected and most of the instances, generated data distribution was within the limit. Situations where, results differed, adjustment of confidence interval could be proposed with justification considering the hyperelestic material properties. Finally, Mooney-Rivlin model was used for data fitting as to further emphasize the results.
| ISBN: | 978-80-7454-965-6 |
|---|
| EAN: | 9788074549656 |
|---|
| Počet stran |
46 stran |
| Datum vydání |
08. 12. 2020 |
| Pořadí vydání |
První |
| Jazyk |
anglický |
| Vazba |
e-kniha - pdf |
| Autor: |
Rohitha Keerthiwansa |
| Nakladatelství |
Univerzita Tomáše Bati ve Zlíně |
| Tématická skupina |
999 - nezařazeno |
| Neprodejná publikace. Publikaci je možné poptávat zde: Volně dostupné na http://hdl.handle.net/10563/45950 |
|---|
Svaz českých knihkupců a nakladatelů