An effective computational tool for parametric studies and identification problems in materials mechanics
Апстракт
Parametric studies and identification problems require to perform repeated analyses, where only a few input parameters are varied among those defining the problem of interest, often associated to complex numerical simulations. In fact, physical phenomena relevant to several practical applications involve coupled material and geometry non-linearities. In these situations, accurate but expensive computations, usually carried out by the finite element method, may be replaced by numerical procedures based on proper orthogonal decomposition combined with radial basis function interpolation. Besides drastically reducing computing times and costs, this approach is capable of retaining the essential features of the considered system responses while filtering most disturbances. These features are illustrated in this paper with specific reference to some elastic-plastic problems. The presented results can however be easily extended to other meaningful engineering situations.
Кључне речи:
Non-linear mechanics / Parametric studies / Identification problems / Proper orthogonal decomposition / Radial basis functionsИзвор:
Computational Mechanics, 2011, 48, 6, 675-687
DOI: 10.1007/s00466-011-0611-8
ISSN: 0178-7675
WoS: 000296784100004
Scopus: 2-s2.0-82955212929
Колекције
Институција/група
VinčaTY - JOUR AU - Bolzon, Gabriella AU - Buljak, Vladimir PY - 2011 UR - https://vinar.vin.bg.ac.rs/handle/123456789/4560 AB - Parametric studies and identification problems require to perform repeated analyses, where only a few input parameters are varied among those defining the problem of interest, often associated to complex numerical simulations. In fact, physical phenomena relevant to several practical applications involve coupled material and geometry non-linearities. In these situations, accurate but expensive computations, usually carried out by the finite element method, may be replaced by numerical procedures based on proper orthogonal decomposition combined with radial basis function interpolation. Besides drastically reducing computing times and costs, this approach is capable of retaining the essential features of the considered system responses while filtering most disturbances. These features are illustrated in this paper with specific reference to some elastic-plastic problems. The presented results can however be easily extended to other meaningful engineering situations. T2 - Computational Mechanics T1 - An effective computational tool for parametric studies and identification problems in materials mechanics VL - 48 IS - 6 SP - 675 EP - 687 DO - 10.1007/s00466-011-0611-8 ER -
@article{ author = "Bolzon, Gabriella and Buljak, Vladimir", year = "2011", abstract = "Parametric studies and identification problems require to perform repeated analyses, where only a few input parameters are varied among those defining the problem of interest, often associated to complex numerical simulations. In fact, physical phenomena relevant to several practical applications involve coupled material and geometry non-linearities. In these situations, accurate but expensive computations, usually carried out by the finite element method, may be replaced by numerical procedures based on proper orthogonal decomposition combined with radial basis function interpolation. Besides drastically reducing computing times and costs, this approach is capable of retaining the essential features of the considered system responses while filtering most disturbances. These features are illustrated in this paper with specific reference to some elastic-plastic problems. The presented results can however be easily extended to other meaningful engineering situations.", journal = "Computational Mechanics", title = "An effective computational tool for parametric studies and identification problems in materials mechanics", volume = "48", number = "6", pages = "675-687", doi = "10.1007/s00466-011-0611-8" }
Bolzon, G.,& Buljak, V.. (2011). An effective computational tool for parametric studies and identification problems in materials mechanics. in Computational Mechanics, 48(6), 675-687. https://doi.org/10.1007/s00466-011-0611-8
Bolzon G, Buljak V. An effective computational tool for parametric studies and identification problems in materials mechanics. in Computational Mechanics. 2011;48(6):675-687. doi:10.1007/s00466-011-0611-8 .
Bolzon, Gabriella, Buljak, Vladimir, "An effective computational tool for parametric studies and identification problems in materials mechanics" in Computational Mechanics, 48, no. 6 (2011):675-687, https://doi.org/10.1007/s00466-011-0611-8 . .