Research @ MIT

1- A New Dynamic Signal-to-Noise Identification Method for Characterization of Fiber Reinforced Textile Composites

During a previous work, a new framework for the inverse identification of material models was developed that is applicable to a variety of engineering problems. The developed multi-objective inverse method was used for the identification of two material models for a typical textile composite under basic deformation modes in quasi-static loading conditions. There is therefore the need for further extension and application of the method to different types of composites under more complex deformation modes, particularly rate/temperature dependent applications. Accordingly, as the main objective of the current research, it is proposed to elaborate on the characterization of composite materials for these applications. To this end, a new dynamic signal-to-noise identification scheme is under development.

Upon the completion of this research, it is expected that the understanding and modeling of non-repeatability in material response, which is a problem that has recently gained significant attention in textile composite literature, will be enhanced. Particularly, it will improve the reliability of rate and temperature dependant constitutive model parameters used in simulation and optimization of forming processes.

2- Application of a new adaptive one-factor-at-a-time design of experiments algorithm in the layout optimization of composite laminates using FEM

The layout optimization of traditional composite laminates has proved to be an unavoidable part of optimal design of sensitive structures such as those in aerospace industries. Two major problems, however, may impede a fast and practical implementation of conventional optimization algorithms for these problems: (a) there are no exact functions available for complex parts and deformation modes to represent the input-output relation analytically, and (b) material and manufacturing noise (such as that due to fiber misalignment in the fabrication stage and that due to inherent variance in measured material properties) can hamper robustness of the designed parts. Furthermore, it is vital to conduct design optimizations with a minimum number of trials in order to cut from experimental/computational costs. To address these issues, a collaborative project is underway with the Robust Design Group at MIT (for the Design of Experiments Part) and Rolls-Royce Canada (for the optimal material selection modeling part).

Upon the completion of this project, it is expected that the layout (both material and stacking sequence) optimization of composite laminates can be performed in a more efficient manner. The proposed technique(s) will also be applicable to advanced 3D textile composites.

Sample Thesis/Industrial Project Collaboration

"A robust multi-criteria approach for experimental optimization of iron oxide nanoparticles for drug delivery", Sharif University of Technology, University of California Davis and MIT, Principal Researcher: M. Mahmoudi (phd candidate at Institute for Nanoscience and Nanotechnology, Sharif University), 2007

"Application of Multiple Criteria Decision making Models for Advanced Material Selection", McGill University and École Polytechnique de Montréal (2005), MIT and Rolls-Royce (present)

"A Multi-objective Non-linear Method for Identification of Rate and Temperature Dependent Metal-Plasticity Models", MIT, McGill University and Rand-North America Inc., 2006

"Equivalent Numerical Model for Honeycomb Subjected to High Speed Impact", Master Thesis by S. Amine, McGill University, Montreal, Canada, 2005

"Modeling and Optimization of Nail Reinforcement in Femur Bones", McGill Orthopedic Research Center, 2005

"Multiple-Criteria Optimization of a Cold Heading Process using Finite Element Analysis and a Taguchi Approach", Master Thesis by C. El-Lahham, McGill University and IVACO, 2004

"Constitutive Models for the Study of Woven Fabrics Under Basic Deformation Modes", Industrial Materials Institute, National Research Council Canada (IMI-NRC), 2002

"Design of Override Gearbox for National Car "ARROW", Co-designers: R. T. Faal and M. Meigunpoori, IUST in collaboration with Mega-Motor Inc., 1997 (selected as the best specialty project of the year by ISME)

Sample Project on Simluation-Based DACE Optimizatin (2005)


Modeling and Optimization of Nail Reinforcement in Femur Bones

The femur or thigh bone is the longest (in length), largest (in volume) and strongest (in mechanical ability to resist deformation) bone of the human body. Osteoporosis in the femur is diagnosed when its strength (often in the neck region) is weakened or it is fractured. The fracture is frequently seen in elderly people, particularly women.

In order to prevent the fracture in a weakened bone, some techniques are used to implant reinforcements into the bone structure. The finite element method has been a powerful tool for the prediction of stress-strain fields in complex structures such as the one in the composite femur bone. In this work, the geometry of a reinforcement is optimized using design and analysis of computer experiments (DACE).

 

* This project was for Dr. Thomas Steffen at Mont-Royal Orthopaedic Research Inc.; I was working in collaboration with the Research Associate, C. El-Lahham.

.

.Computer Skills

Abaqus (user defined material modeling, object oriented python scripting), SolidWorks, Catia, Autocad, Matlab, Mathematica, LabView, Visual Basic, Fortran, C++, MOWR (multi-objective weighted regression code, developed during Ph.D.), Gearbox Design Package (developed during B.Sc. & M.Sc.)