Mohammad Zamani

About Me

I am a Research Fellow at the High Performance Computing Lab, School of Civil Engineering, University of Tehran, under the supervision of Professor Soheil Mohammadi.

I specialize in developing innovative solutions for complex engineering challenges. My research focuses on:

Computational Mechanics: Finite element analysis, multiphysics modeling, and numerical simulations
Machine Learning in Engineering: Deep learning, reinforcement learning, and AI-driven optimization
Multiscale Modeling & Optimization: Topology optimization, composite materials, and mechanical metamaterials
Advanced Computational Methods: High-performance computing, parallel processing, and adaptive algorithms

Location

CHPC Lab, University of Tehran

Tehran, Iran

View on Map

Email

mail.zamani.m@gmail.com

Education

M.Sc. in Structural Engineering

University of Tehran

2019 - 2022

School of Civil Engineering - High-Performance Computing Laboratory
Thesis: Mathematical Modeling of Bone Fracture Healing
Developed a novel computational framework using finite element methods to simulate and analyze the complex biological processes involved in bone fracture healing, incorporating coupled reaction-diffusion equations.
GPA: 16.80/20.0 (Upper Half of Class)
Key Courses: (Non)linear FEM, Continuum Mechanics, Multiscale Methods, Optimization, ML and RL
Research Focus: Computational Mechanics, Multiscale Modeling, Machine Learning, Biomechanics

B.Sc. in Civil Engineering

Hekmat University

2014 - 2017

School of Civil Engineering
GPA: 15.62/20.0 (Upper Third of Class)
Key Courses: Structural Analysis, Mechanics of Materials, Numerical Methods, Programming
Senior Project: Design and Analysis of a Multi-Story Building

Research Experience

Graduate Research Assistant

2021 - Present

University of Tehran - HPC Lab

Computational Biomechanics

Developed a novel FEM framework for tissue vascularization simulation
Solved coupled reaction-diffusion equations numerically

Machine Learning in Engineering

Led comparative analysis of ML methods for engineering datasets
Developed deep learning models for material property prediction
Implemented reinforcement learning for structural optimization

Multiscale Modeling & Optimization

Improved homogenization methods for composite materials
Developed topology optimization algorithms for lightweight structures
Created inverse design methods for mechanical metamaterials

Graduate Research Projects

2019 - 2022

University of Tehran

Advanced Computational Methods

Implemented adaptive FEM solvers in MATLAB and Python
Developed meshless methods for complex geometries
Created parallel computing algorithms for large-scale simulations

Materials Science Applications

Applied multiscale modeling to composite materials
Developed micromechanics models for material behavior
Implemented statistical mechanics approaches for material properties

Publications

Journal Papers

The Impact of Data Splitting Methods on Machine Learning Models: A Case Study in Predicting the Concrete Workability

Machine Learning for Computational Science and Engineering, 2025

DOI: 10.1007/s44379-025-00021-3

A structured evaluation framework for assessing concrete workability in a more efficient and sustainable manner.
Consistency in data splitting to ensure reliable and reproducible model assessment.
Nested cross-validation to minimize sampling effects and improve evaluation robustness.
Deep neural networks (DNNs) for enhancing accuracy in predicting concrete properties from imbalanced datasets.
Multi-output DNNs and transfer learning to exploit shared property correlations for better flow prediction.

Finite Element Solution of Coupled Multiphysics Reaction-Diffusion Equations for Fracture Healing in Hard Biological Tissues

Computers in Biology and Medicine, 2024

DOI: 10.1016/j.compbiomed.2024.108829

Finite element solution of the reaction-diffusion equations governing fracture healing in hard tissues.
Weak formulation to enhance stability for complex domains, coarser meshes, and accurate boundary conditions.
Captures various stages of fracture healing, e.g., soft and hard callus formation, and endochondral ossification.
Predictions demonstrate coherence with available reference experimental and numerical data.

Conference Papers

3D Multiscale Topology Optimization for Conceptual Design of a Quadrotor Aerial Taxi

The 33th Annual International Conference of Iranian Society of Mechanical Engineers, 2025

DOI: 10.1234/isme.2025.12345

Developed a computational framework for 3D concurrent topology optimization of multiscale composite structures.
Combined modified SIMP method with asymptotic homogenization for effective material properties.
Implemented 3D eight-node hexahedral elements at both macro and micro scales.
Achieved optimal combination of lightness, strength and mechanical stability for aerial taxi design.
Demonstrated significant impact of asymptotic homogenization in composite design accuracy.

Inverse Design of New Mechanical Metamaterial for Base Isolator

The 33th Annual International Conference of Iranian Society of Mechanical Engineers, 2025

DOI: 10.1234/isme.2025.4321

Developed topology optimization framework for mechanical metamaterials with high bulk-to-shear modulus ratio.
Introduced novel filtering function maintaining connectivity and symmetry in optimization.
Implemented 3D inverse homogenization framework with energy-based property computation.
Achieved optimal metamaterial design for seismic base isolation applications.
Demonstrated rational design approach for metamaterials with tunable elastic properties.

Book Chapter

Biomechanics of Hard Tissues (Chapter 6) in Multiscale Biomechanics

Ed. S. Mohammadi, Wiley, 2023

DOI: 10.1002/9781119033714.ch6

Analysis of macro and micro structures in hard tissue architecture.
Implementation of numerical simulations.
Investigation of healing processes through governing equations and numerical methods.