Research & Publications

Publications

Compressive Strength Prediction for Industrial Waste-Based SCC Using Artificial Neural Network

Md Akram Hossain, Islam, G. M. S., & Amit Mallick. (2022). Compressive Strength Prediction for Industrial Waste-Based SCC Using Artificial Neural Network. Journal of the Civil Engineering Forum, 9(1), 11-26. https://doi.org/10.22146/jcef.4094

View Publication

Key Metrics & Methodology:

Architecture: 9-25-1 Back-Propagation Neural Network (BPNN) via Levenberg-Marquardt algorithm
Performance: R² = 0.9835 | RMSE = 2.4503 | MAPE = 0.0347
Dataset: Evaluated global database of 354 mix proportions incorporating industrial waste (fly ash & silica fume)

Ongoing Research

Improvement of Concrete Breakout Strength of Anchor bolt

In collaboration with Arman Engineering Ltd. & CUET.

Methodology & Focus:
Non-linear finite element modeling (Abaqus) and experimental testing of post-installed anchors near thin slab edges (125 mm thickness).

Key Innovations:
Evaluating supplementary hairpin reinforcement detailing and Fiber-Reinforced Concrete (FRC) matrix modifications to suppress concrete breakout failure cones and reduce edge-distance penalties.

AI Integration:
Developing machine learning surrogate models to predict failure loads under combined shear-tension interaction.

Design Strength Improvement of Structural Silicone

In collaboration with Arman Engineering Ltd. & CUET.

Methodology & Focus:
Hyperelastic material characterization (Mooney-Rivlin / Ogden constitutive models) combined with non-linear FEA simulation.

Key Innovations:
Investigating stress distribution profiles across structural silicone sealants under high wind-load cycling to optimize joint geometries and material usage.