Civil Engineering


First Advisor

Professor Chung R. Song

Date of this Version

Summer 7-2-2018


Koocheki, Kianoosh. (2018). Artificial Neural Network and Finite Element Modeling of Nanoindentation Tests on Silica (Master's Thesis). University of Nebraska-Lincoln Digital Commons


A THESIS Presented to the Faculty of The Graduate College at the University of Nebraska In Partial Fulfillment of Requirements For the Degree of Master of Science, Major: Civil Engineering, Under the Supervision of Professor Chung R. Song. Lincoln, Nebraska: June 2018

Copyright (c) 2018 Kianoosh Koocheki.


Two major forms of Silica include the crystalline form named Quartz which consist of the sand grains in nature, and amorphous form named Silica Glass or Fused Silica which is commonly known as glass. Fused Silica is an amorphous crystal that can show plastic behavior at micro-scale despite its brittle behavior in large scales. Due to the amorphous and ductile nature of Fused Silica, this behavior may not be explained well using the traditional dislocation-based mechanism of plasticity for crystalline solids. The crystal plasticity happens due to shear stress and stored energy in the material as dislocations which does not change the volume. In amorphous Fused Silica however, the permanent deformation is mainly caused by densification of the material under localized loading in addition to plastic flow caused by shear stress. This behavior is particularly true in the case of nanoindentation testing. Due to this densifying behavior, modeling the material using constitutive models such as Drucker-Prager/Cap can be quite helpful to further expand the model parameters to be used for geomaterials. Nanoindentation tests were performed on Fused Silica and Quartz samples and Finite Element Method (FEM) was used to further investigate the effect of different constitutive model parameters on material behavior. It was observed that, by implementing volumetric hardening in constitutive models, the FEM results were in better agreement with experimental results in case of both Fused Silica and sand grains. In the second part of the study Artificial Neural Network (ANN) models were used to predict nanoindentation test results for different material parameters as well as indenter shape and geometry. ANN models were trained using FEM results and experimental test results and verified using the reminder of the data. Trained models were then used to study of different scenarios that were not analyzed using FEM or experiments.

Advisor: Chung R. Song