Constitutive models and statistical analysis of the short-term tensile response of geosynthetics after damage
Conference Paper
Overview
Research
Identity
Additional Document Info
View All
Overview
abstract
Geosynthetic is a generic name given to planar products, mostly composed of thermoplastic polymers, and used in contact with soil, rock or with any other material as part of a construction system [1]. Geosynthetics have several functions and may perform more than one simultaneously, such as soil reinforcement, stabilization of steep slopes, filtration, drainage, fluid barrier, erosion control and coastal protection [2].
The objective of this research was to analyse the short-term tensile response of three geosynthetics using the procedures described by [3], and to apply constitutive equations to represent the nonlinear behaviour of the materials. Data on specimens of a nonwoven polypropylene geotextile (GTX), a woven polyester geogrid (GGR) and a reinforcement polyester geocomposite (GCR) were analysed. Some specimens of each material were submitted to mechanical damage [4], abrasion [5], and mechanical damage followed by abrasion. Nonlinear regressions of the experimental data were performed to fit the load-strain curves to a hyperbolic-based equation depending on the tensile response of the geosynthetic: type A (GTX) or type B (GGR and GCR) [6]. For each geosynthetic, the results of damaged specimens were statistically compared to those of the undamaged specimens to observe the influence of the induced damage on the tensile behaviour of the material. Experimental data were statistically compared with those fitted by the constitutive models to verify if the tensile properties were properly estimated – namely the secant stiffness for 2% strain, the ultimate tensile strength, and the strain at maximum load.
For the GTX, significant changes in tensile properties occurred only after mechanical damage followed by abrasion. For the GGR and the GCR, abrasion was the predominant damage due to considerable changes in the tensile properties and the shape of the load-strain curves. In general, the hyperbolic-based models presented good approximation of the empirical data. Curves for damaged materials were plotted using undamaged model parameters and applying adjustment coefficients and reduction factors allowing for damage, in which the goodness-of- fit was considered promising.
