Research

Direct Ink Writing of Advanced Nanocomposite Rubber Compounds

Application of Machine Learning in Design

Goal: Use machine learning algorithms as useful tools to solve mechanical engineering problems.

Sample Project: Smart Flexible Sensors

We developed a robust and intelligent sensor capable of detecting pressure magnitude, location, and the size of the contact region, even with random noise and uncertainty.

Sensor Design: Interdigitated copper electrodes that provide unique readings for different loads applied on the sensitive layer.

Deep Neural Network models were built for accurate and robust prediction of pressure distribution (magnitude, location, and the size of contact region).

 

 

Design and Additive Manufacturing of Architectured Materials

Goal: Design and 3D print architectured polymer nanocomposites for different multifunctional applications.

The evolution of additive manufacturing techniques has enabled design freedom that can be exploited for tailoring the behavior of materials for various applications.

Sample Project: Design for flexible skin sensors 

Kirigami and bio-inspired structures are designed and 3D printed using piezoresistive flexible materials to study the effect structures on the electromechanical behavior of skin sensors.

Design for Skin Sensors 

 

Sample Project: Design for maximum auxeticity and flexibility

Auxetic structures exhibit unique mechanical behavior and have a negative Poisson’s ratio which means under tension they become thicker perpendicular to the applied load. This unique behavior can be exploited in many applications.

Design for Flexibility and Auxeticity

 

Advanced Manufacturing Processes Simulation

Goal: Find the optimum processing parameters to maximize the quality of the manufactured part

Sample Project: Simulation of Automated Fiber Placement Process

A Finite Element Model was used to simulate the evolution of temperature and pressure profiles in the composite layers during the automated fiber placement process. The numerical results along with intimate contact and autohesion models are used to predict the evolution of bonding (DoB) between the composite layers and find the optimized processing parameters in the automated fiber placement that maximize the bonding between the composite layers.

Finite element simulation of automated fiber placement (the roller and heating source are models as moving applied pressure and heat input)

 

Adhesion and Friction Behavior of Viscoelastic Materials

Goal: Develop a Physics-based Multiscale Model to Predict the Adhesion and Friction Coefficient of Viscoelastic Materials.

Viscoelastic materials exhibit both viscous and elastic characteristics and behave both like a liquid and a solid. Friction and adhesion of these materials are of significant importance in many practical applications. A physics-based multiscale model allows us to predict and tune the behavior of these materials for different applications.

Sample Project: Friction Prediction in Rubber Compounds

We used a multiscale physics model based on surface roughness power spectral density of contact surfaces and the time-dependent behavior of rubber to predict friction coefficient in tire tread compounds.

Left) 3D scan of the surface profile of asphalt pavement. Right) Line scan of surface profile with height distribution.

 

 

 

 

 

 

 

Left) Surface roughness power spectral density Right) Theoretical friction coefficient prediction and experimental results