SOLVE - is a research hub devoted to advancing the science and practice of performance-driven structural design.

 

We work at the intersection of structural optimization, computational mechanics, experimental testing, and machine learning to deliver structures that are lighter, safer, and more sustainable - without compromising serviceability or constructability. Our mission is to translate rigorous research into viable engineering: solutions that stand up scientifically, scale in the field, and remain economical over their life cycle.

 

Modern infrastructure faces competing demands: higher performance with lower embodied carbon; resilience under multi-hazard actions; and faster, more reliable delivery. SOLVE approaches these challenges through shape and performance optimization coupled with high-fidelity finite element modeling (ABAQUS implicit/explicit) and targeted laboratory validation.

 

Our in-house, gradient-free optimization frameworks - GSO (Gradient-less Shape Optimization) and SSCO (Simultaneous Shape and Cable Optimization), navigate high-dimensional, multi-criteria design spaces to identify Pareto-optimal solutions. Across beams, plates with cut-outs, sandwich and prestressed elements, these methods have delivered 15–25% weight reduction while maintaining stiffness and serviceability, demonstrating advantages over traditional gradient-based techniques.

 

A signature thrust of SOLVE is advanced composite sections, like Concrete-Filled Steel Tubular (CFST) configuration. Through coordinated experiments and numerical simulations, we optimize confinement, hollow ratios, tube geometry, and material pairing to elevate axial and lateral performance. These studies have shown substantial gains in load capacity and ductility compared to conventional reinforced concrete (RC) solutions.

 

Complementary research spans blast/impact resistance, seismic response, frame-action behavior, and residual strength after extreme events. Our methodology blends physics-based modeling with data-driven intelligence. Parametric FEM models are automated for rapid geometry generation, meshing, material calibration, and result extraction; these datasets feed ML pipelines for sensitivity analysis, surrogate modeling, and predictive strength/ductility estimations. This digital workflow accelerates design space exploration and supports decision-ready insights for practicing engineers and asset owners.

 

Sustainability is woven through our work. We investigate low-carbon infill materials (e.g., geopolymer mixes, enhanced recycled-aggregate concrete, cement-reduced UHPC) and quantify trade-offs using performance metrics alongside durability, energy absorption, and residual capacity. Our objectives align with SDG 9, 11, 12, and 13, targeting material efficiency, extended service life, and reduced lifecycle impacts. Where appropriate, we connect optimization with retrofit strategies (e.g., FRP strengthening) to upgrade existing assets rather than replace them.

 

SOLVE bridges lab and field. Team members have contributed to 70+ large-scale projects - from institutional buildings and elevated roads to bridge assessments and rehabilitation - bringing practical constraints into the research loop. This feedback from real projects informs our assumptions, validates models, and shapes design guidance that practitioners can trust.

 

We collaborate with international researchers and industry partners, publish in leading journals, and contribute to the community through peer review, workshops, and open dialogues. As the lab grows, we aim to release selective tools, datasets, and benchmark problems to help the community evaluate new algorithms and design concepts on common ground.