Seeking a PhD position in Structural Engineering · Fall 2027

Md Hasibul Islam(Nabil)

Civil Engineering (Structural) Undergraduate, BUET

I study FRP-reinforced and lightweight concrete, and use machine learning to make structural engineering faster and more accurate.

Dhaka, Bangladesh

Portrait of Md Hasibul Islam

About

Structural engineering, with code

I'm a final-year Civil Engineering undergraduate at Bangladesh University of Engineering and Technology (BUET), majoring in structural engineering with a minor in geotechnical engineering.

My research is about FRP-reinforced and lightweight concrete. I also build software: Python and MATLAB tools for structural analysis, and machine-learning models for problems in civil engineering.

Research interests

  • FRP composites
  • Sustainable & lightweight concrete
  • Bond and anchorage
  • Finite element modeling
  • Machine learning for structures

Research

Current work & highlights

Undergraduate thesis

Bond-slip behavior of GFRP bars in recycled-pumice lightweight concrete

Supervisor: Prof. A. F. M. Saiful Amin

The concrete uses pumice, a waste product of Bangladesh's denim-washing industry, as fine aggregate and recycled demolition concrete as coarse aggregate. ACI CODE-440.11-22 leaves lightweight concrete out of its GFRP design scope, and no one has yet characterized bond-slip behavior in this material system.

  • ASTM D7913 pull-out tests with LVDT slip measurement
  • 10–50 MPa concrete mixes, 8–20 mm bar diameters
  • Abaqus finite element models to validate the experiments

Competition · Computational Analysis Lead

ACI FRP Composites Competition 2026

ACI Spring Convention · Rosemont, Chicago · March 2026

Our eight-member BUET team designed, built and tested a 1 m concrete beam reinforced with GFRP bars, made with pumice-based lightweight concrete. Each team had to predict its beam's ultimate load and failure mode in advance and achieve the lowest cost-to-load ratio.

  • Wrote a Python program that generated the beam's stress trajectories. We used it to place the FRP bars to resist both moment and shear at minimum cost.
  • Worked in the lab on concrete mix design and beam forming.
Official results (PDF)
  • 4th Place Award (tied on points for 3rd) among 43 teams from 11 countries
  • 1st in load-prediction accuracy: predicted vs. measured ultimate load within 0.47%
  • Highest ultimate load of any ranked team

Software

Apps I've built

Beam Calc

Beam, frame & truss solver for Android

  • 5K+ downloads
  • 5.0★ rating

A structural analysis app for civil engineering students and engineers. Model a beam, 2D frame or truss and get shear force, bending moment, slope and deflection diagrams and member forces instantly.

  • Kotlin
  • Jetpack Compose
  • Direct stiffness method
Get it on Google Play
  • Indeterminate beams and frames solved by the direct stiffness method; determinate beams by statics
  • Supports, internal hinges, point/UDL/triangular/trapezoidal loads, and moments
  • Variable E and I, support settlement and rotation
  • Influence lines, degree of indeterminacy, and detection of unstable layouts

Publications

Conference papers

  1. Revolutionizing Concrete Mix Design with Neural Networks: A Machine Learning Approach

    M. H. Islam, M. R. H. Sarkar, M. M. Rahman, S. Islam, M. F. Alam

    8th ICERIE 2025, SUST · Poster presentation· Apr 2025First author

    Feedforward neural network trained in MATLAB on 500,000 synthetic ACI and BS mix designs (R² = 0.97). It gives continuous mix proportions and a sensitivity analysis without chart lookups.

  2. Automated Liquid Limit Determination Using Video and Acoustic Analysis of Casagrande's Test

    M. H. Maruf, M. A. Rahman, M. H. Islam, S. Islam, M. F. Alam

    ICCEI 2025, RUET· Dec 2025Presented

    Automates the Casagrande test with a smartphone. Blows are counted from audio peaks and groove closure is measured frame by frame (CIE Lab segmentation, K-means). Liquid limits for three clays came within 1–11% of the manual method.

  3. Image Processing-Based Automated Detection and Counting of Coliform Colonies on Petri Dishes for Water Quality Analysis

    M. H. Maruf, M. H. Islam, H. Mahmud, S. K. Sunny

    ICCEI 2025, RUET· Dec 2025

    U-Net segmentation model trained on 58,000 synthetic Petri-dish images rendered in Blender. It counts coliform colonies from ordinary smartphone photos within 1–10% of manual counts.

Background

Experience & education

Experience

  • Director (Research & Development)

    ACI BUET Student Chapter

    Jul 2026 – Present

    • Executive panel member of the American Concrete Institute's BUET student chapter
    • Led the development of the chapter's website
    • Organizes chapter events and industry site visits
  • Computational Analysis Lead

    ACI FRP Composites Competition · BUET

    Oct 2025 – Mar 2026

    • Built the stress-trajectory tool used to place the FRP reinforcement
    • Concrete mix design and beam forming in the lab

Education

  • B.Sc. in Civil Engineering

    Bangladesh University of Engineering and Technology (BUET)

    2022 – 2027 (expected)

    • Major: Structural Engineering · Minor: Geotechnical Engineering
  • Higher Secondary Certificate, Science

    Notre Dame College, Dhaka

    2019 – 2021

Skills

Tools I work with

Engineering

  • Structural Engineering
  • Reinforced Concrete
  • Fiber Reinforced Polymer (FRP)
  • Concrete Mix Design
  • Finite Element Analysis

Software & Code

  • Python
  • MATLAB
  • Abaqus
  • C++
  • Android (Kotlin, Jetpack Compose)
  • Web Development

Machine Learning

  • Neural Networks
  • Computer Vision
  • Image Segmentation (U-Net)
  • Synthetic Data (Blender)

Contact

Let's talk research

I'm looking for PhD positions in structural engineering starting Fall 2027, and I'm always happy to talk about FRP, concrete or ML in civil engineering.