Research Topics
-
- Mechanics and Multi-physics of Dissipative Heterogeneous Materials
- Physics-based Machine Learning to Predict Damage
- Sensitivity Analysis of Failure in Heterogeneous Materials
- Cracks Initiation and Propagation based on Sensitivity Analysis and Normality Tests
- Sustainability and Resiliency of Infrastructure Materials
Mechanics and Multi-physics of Dissipative Heterogeneous Materials
We focus on the mechanics and multi-physics of heterogeneous materials with rate-and temperature-dependent matrices. Our goal is to understand and characterize the effects of environmental factors such as temperature, oxygen, moisture, and chloride on the response of such composite materials. We develop thermo-hydro-chemo-mechanical constitutive equations based on theories of thermodynamics and physics. Asphalt concrete and elastomers with different fillers are two examples of our current materials of interest.
Current student: Aimane Najmeddine, Marwa Yacouti
Sponsor: NSF
Physics-based Machine Learning to Predict Full-Field Inelastic Responses and Damage
Our goal is to understand and characterize damage in different classes of composite materials based on physics-based machine learning approaches. We integrate physics and mechanics laws within the training networks to obtain more accurate predictions of materials responses. FE simulations are used to produce a sufficient amount of results to train and verify the algorithms.
Current student: Reza Sepasdar
Sponsor: AFOSR, Virginia Tech
Sensitivity Analysis of Failure in Heterogeneous Materials
The microstructural properties in fiber-reinforced composites are not distributed uniformly due to manufacturing processes. Thus, it is of utmost importance to characterize the sensitivity of macrostructural responses to random microstructural properties. We develop analytical sensitivity computations using the direct approach to derive derivatives of the transverse failure response with respect to random fiber/matrix interfacial properties. The obtained results are capable of predicting the initiation and propagation of transverse cracks.
Current student: Luis Hernandez
Sponsor: AFOSR, Virginia Tech
Cracks Initiation and Propagation based on Sensitivity Analysis and Normality Tests
The sensitivity analyses results provide valuable information on the sensitivity of the response with respect to local variations in microstructural properties. The results will help us to understand (1) which fibers are the most effective in the overall response of the composites, (2) when and where cracks initiate, and (3) what is the distribution of extreme values (properties with maximum or minimum sensitivities). In the end, we will be able to predict the probability distribution of composites response based on the available information on the probability distribution of fibers location, and mechanical properties of the constituents without running several simulations. These understandings will provide a tool to design composites which are insensitive to material properties and geometrical variations.
Sustainability and Resiliency of Infrastructure Materials
Pavement properties affect vehicle fuel consumption and therefore induce extra emission through different phenomena. This effect needs to be considered in pavement design. We developed a mechanics-based approach to incorporate the effect of pavement structure on vehicle fuel consumption. The obtained model was implemented in a pavement Life Cycle Assessment (LCA) too.