Valdes et al 2021a - Revision history

Scipediacontent at 15:51, 26 June 2021

2021-06-26T15:51:20Z

Scipediacontent: Scipediacontent moved page Draft Content 452585662 to Valdes et al 2021a

2021-06-26T15:48:05Z

Scipediacontent moved page Draft Content 452585662 to Valdes et al 2021a

Scipediacontent: Created page with "== Abstract == Data-driven Identification (DDI) is a technique that allows to estimate the stresses of a sample and the behavior of the material solely by the use of strain i..."

2021-06-26T15:48:03Z

Created page with "== Abstract == Data-driven Identification (DDI) is a technique that allows to estimate the stresses of a sample and the behavior of the material solely by the use of strain i..."

New page

== Abstract ==

Data-driven Identification (DDI) is a technique that allows to estimate the stresses of a sample and the behavior of the material solely by the use of strain information, avoiding the bias imposed by an empirical constitutive model. In this work, we extend the applicability of DDI from elasticity to linear-viscoelastic materials by extending the dimensionality of the problem. Rather than estimating the state of the elements considering an instantaneous value of strain-stress, we include the strain history of the sample in order to account for the viscosity effect. We also combine the method with data analysis techniques such as Kernel Principal Component Analysis to improve the estimation of stresses. Preliminary results in modeled samples show a clear improvement on the estimation of stresses when compared against the original formulation of the algorithm, allowing us to obtain results in cases where the original DDI fails to do so.

== Full document ==
<pdf>Media:Draft_Content_452585662A_IDC6_135.pdf</pdf>

@@ Line 1: / Line 1: @@
-== Abstract ==
+== Summary ==
 Data-driven Identification (DDI) is a technique that allows to estimate the stresses of a sample and the behavior of the material solely by the use of strain information, avoiding the bias imposed by an empirical constitutive model. In this work, we extend the applicability of DDI from elasticity to linear-viscoelastic materials by extending the dimensionality of the problem. Rather than estimating the state of the elements considering an instantaneous value of strain-stress, we include the strain history of the sample in order to account for the viscosity effect. We also combine the method with data analysis techniques such as Kernel Principal Component Analysis to improve the estimation of stresses. Preliminary results in modeled samples show a clear improvement on the estimation of stresses when compared against the original formulation of the algorithm, allowing us to obtain results in cases where the original DDI fails to do so.
-== Full document ==
+== Document ==
 <pdf>Media:Draft_Content_452585662A_IDC6_135.pdf</pdf>

← Older revision	Revision as of 15:48, 26 June 2021
(No difference)