This project develops simulation software for heat-pump water-heater performance evaluation, single-phase cold-plate data-center component sizing, waterfall-PUE analysis, and system-level performance calculation under different design and operating conditions. The tool supports component-level and system-level studies of liquid-cooled data centers with integrated waste-heat recovery via heat-pump water heaters.
Built component sizing and performance models for single-phase cold plates.
Implemented waterfall-PUE evaluation and overall system performance calculations.
Coupled the data-center cooling loop to a heat-pump water-heater for waste-heat recovery analysis.
A design and analysis platform for liquid-cooled data centers with integrated heat recovery via heat-pump water heaters.
This project develops distributed and discretized models for single-phase and two-phase plate heat exchangers, capturing heat-transfer prediction, pressure-loss calculation, and flow-maldistribution analysis across the plate stack. The simulation framework supports sizing and off-design performance evaluation for industrial heat-exchanger design.
Built distributed models for single-phase and two-phase plate heat exchangers.
Captured pressure-loss and flow-maldistribution effects across the plate stack.
Implemented sizing and off-design performance routines for industrial use.
A sizing and performance-evaluation tool for plate heat exchangers in industrial cooling applications.
This project develops simulation software for solid-oxide fuel-cell (SOFC) heat-recovery applications, including phase-change material (PCM) thermal storage, organic Rankine cycle (ORC) integration, and double-effect absorption chiller coupling. The framework supports system-level analysis of SOFC waste-heat utilization across power, heating, and cooling outputs.
Developed integrated models for SOFC heat recovery with PCM storage, ORC, and absorption chillers.
Built a system-level simulation framework for trigeneration analysis.
Enabled performance evaluation across multiple operating scenarios.
A simulation tool for evaluating SOFC-based trigeneration systems combining electricity, heat, and cooling.
This project develops Modelica-based dynamic simulation models for heat-recovery data-center and integrated energy-system applications under part-load operating conditions. The system integrates a cascade of high-temperature steam electrolysis (HTSE), organic Rankine cycle (ORC), and multi-effect distillation (MED) subsystems with supervisory and local controllers designed for stable operation across a wide range of part-load scenarios. The work extends the NHES Modelica library with project-specific components and includes numerical commissioning to ensure robust transient behavior.
Developed dynamic Modelica models for the HTSE / ORC / MED cascade.
Designed supervisory and local control strategies for part-load operation.
Performed numerical commissioning and stability tuning for transient simulations.
Added project-specific components to the NHES Modelica library.
A dynamic-simulation framework for evaluating integrated energy systems combining heat recovery with data-center cooling demand.
This project focused on the modeling and control of vapor-injection heat pump water heaters using Modelica. A physics-based, acausal dynamic model was developed, formulated as a system of differential-algebraic equations (DAEs). The evaporator and condenser were modeled using the moving boundary method, capturing two-phase flow dynamics under transient conditions. Smoothing functions were applied to transition regions to ensure differentiability and stable numerical integration. The system incorporates two Electronic Expansion Valves (EEVs)—one for the main refrigeration cycle and another for the vapor-injection loop. Separate control strategies were designed for each: PI controllers for maintaining target operating conditions (e.g., superheat or injection pressure) and extremum-seeking controllers for optimizing performance based on system efficiency or heat output. Controllers were evaluated within the transient, physics-based simulation environment under varying load and weather conditions.
Developed a transient, physics-based model in Modelica using an acausal DAE formulation.
Implemented moving boundary models for evaporator and condenser dynamics.
Applied smoothing functions to handle phase-region transitions in the numerical model.
Designed control logic for two EEVs, one in the main loop and one in the injection loop.
Applied PI and extremum-seeking control strategies for coordinated operation of both EEVs.
Built a modular simulation environment for performance evaluation under dynamic conditions.
Delivered a dynamic simulation framework for vapor-injection heat-pump water-heater systems with multi-evaporator capability, including PI-controller tuning and extremum-seeking control design; model validation showed mass- and energy-balance errors within approximately ±1–2%.
This project focused on the development, experimental evaluation, and dynamic simulation of a liquid desiccant dehumidification system using the ionic liquid CreCOPlus® 5100 (Evonik) under the supervision of Prof. Chi-Chuan Wang. The system includes internally-cooled and internally-heated configurations and is intended for integration with energy-efficient heat pumps in net-zero energy buildings. Experiments were carried out to study key thermal and mass transfer behaviors, with a focus on response time under varying operating conditions. A physics-based dynamic simulation model was developed and validated using Julia’s DifferentialEquations.jl package under the supervision of Dr. Chris Rackauckas (JuliaHub/MIT). We also compared model performance and runtime across Julia, MATLAB, and SciPy implementations. The project presents a framework and set of guidelines for evaluating system feasibility under specific scenarios. While the current focus is on standalone operation, the methodology is generalizable and applicable to configurations involving integration with systems such as DOAS, All-Air systems, or VRF systems.
Designed and built experimental setups for both the dehumidifier and regenerator units
Developed detailed steady-state models capturing coupled heat and mass transfer in ionic liquid systems
Validated simulation models using experimental data and incorporated them into sensitivity analysis workflows
Simulated system performance across a range of inlet air conditions, solution concentrations, and flow rates
Defined and analyzed the system’s feasible region to assess operational limits and design viability
Delivered a high-efficiency Julia-based simulation code for feasible-region analysis, achieving approximately 40× faster runtime than MATLAB and 800× faster runtime than SciPy, while identifying the feasible operating range of ionic-liquid mass fraction and solution temperature for dehumidification performance.
This project supported the development of hybrid Physics-Informed Neural Network (PINN) modeling workflows for two-phase immersion cooling of electronic packages, blending PINN physics-based losses with a sparse subset of CFD-derived labeled points to capture boiling dynamics and heat transport under varying load conditions.
Contributed to hybrid PINN model development supplementing physics-based losses with ~20% of CFD-labeled fluid-domain points per snapshot for two-phase immersion-cooling problems.
Implemented a VOF-aware spatial sampling strategy that blends a Sobol space-filling rank with an interface-strength rank to preserve liquid–vapor interface resolution under sparse labeling.
Supported training pipelines and validation against finite-volume CFD data.
A submitted journal manuscript (under minor revisions) and a delivered hybrid-PINN modeling pipeline that achieves wall-superheat prediction within ~2.3% using only ~20% of CFD labels per snapshot, with field-level RMSE on the order of 10⁻³ for vapor volume fraction and velocity magnitude.
This project supported experimental testing, verification, and PINN-based prediction for air-cooled data-center racks, including study of arrangement-dependent recirculation in cold-aisle-contained configurations with partially populated racks.
Supported experimental data analysis and verification of rack airflow patterns.
Contributed to PINN-based prediction models for thermal behavior of partially populated racks.
A submitted journal manuscript and a PINN-based prediction tool for arrangement-dependent thermal effects in air-cooled data centers.
This project supported hybrid PINN workflows for water-side and air-side physical modeling of data centers, integrating airflow and thermal predictions in a unified neural-network framework.
Contributed to development of hybrid PINN models coupling water-side and air-side data-center physics.
Supported model validation against experimental datasets.
An accepted journal manuscript and a PINN-based modeling framework for coupled airflow-thermal analysis of data centers.
This project involved the development of a hybrid simulation framework for vapor-compression systems, combining gray-box and physics-based modeling approaches. Curve-based gray-box models were created for the compressor and electronic expansion valve (EEV), offering computational efficiency and ease of calibration. For the evaporator and condenser, physics-based models were developed using the moving boundary method to capture transient two-phase flow and heat transfer behavior. The simulation environment supports systems with multiple evaporators operating under varying conditions and was implemented in Python.
Led the project and coordinated development efforts with the industry partner's collaborators.
Developed gray-box models for the compressor and EEV using curve-fitting techniques.
Implemented physics-based moving boundary models for condensers and evaporators.
Enabled simulation of multi-evaporator configurations within a unified framework.
Designed a modular, extensible architecture to support future model expansion.
Validated system behavior against performance data and representative operating conditions.
The completed models were integrated into the industry partner’s internal simulation platform under confidentiality agreements, providing a flexible foundation for further component development and system-level studies.
This project focused on developing a simulation tool for a hybrid cooling system that combines an evaporative condenser with a cooling tower. The system improves water-side heat rejection by spraying water over the condenser tubes, which absorbs heat before passing through a cooling tower for recirculation. The simulation models the thermodynamic and psychrometric interactions between airflow, water flow, the condenser, and the cooling tower under a crossflow configuration.
Designed and implemented the system model and software architecture for the hybrid cooling setup.
Developed a modular simulation engine to capture evaporative heat transfer and air–water interactions.
Built Python-based solvers with configurable inputs for flow rates, fan characteristics, and ambient conditions.
Simulated spray behavior, temperature reduction across the cooling tower bed, and water recirculation performance.
Provided a framework to support design evaluation and performance analysis under varying environmental conditions.
The software was delivered to the industry partner under confidentiality agreements and is intended for internal use in system design analysis and future development.
This project continued the earlier work on evaporative condenser modeling, focusing on configurations with staggered circular copper tubes. A Python-based solver was implemented using the moving boundary method. The software included a GUI developed with PySide, and followed a consistent software structure using common object-oriented design patterns. CoolProp was used to support multiple refrigerants.
Implemented a simulation model for copper-tube condensers using the moving boundary approach.
Developed a GUI in PySide consistent with the earlier project phase.
Maintained modular and organized software architecture using standard design patterns.
Enabled refrigerant flexibility by integrating CoolProp.
The final software was submitted to the industry partner under confidentiality agreements for internal use.
This project focused on developing an initial simulation tool for evaporative condensers using inline aluminum flat tubes. The tool was based on the moving boundary method and implemented in Python. A graphical user interface (GUI) was created using PySide, and the software followed established object-oriented design patterns such as MVC, Abstract Factory, Observer, and Strategy. CoolProp was integrated to provide refrigerant property data.
Developed a Python-based simulation model for inline flat-tube condensers.
Created a user interface using PySide with a modular code structure.
Applied standard object-oriented patterns to improve code organization.
Integrated CoolProp for refrigerant thermophysical property support.
Provided a codebase that supported further development in later project phases.
The software was delivered to the industry partner under confidentiality agreements and used as a foundation for follow-up work.
This project focused on creating a simulation and design environment for solar water heating systems, aligned with international thermal performance standards. Although not governed by a specific national code, the tool was customized for local Iranian conditions and designed to be accessible for engineers and technicians operating in diverse climatic regions.
Performed comprehensive assessments of technical, environmental, and geographic factors.
Developed a regionally adapted simulation tool for solar water heating system design.
Applied engineering modeling methods consistent with international performance standards.
Designed intuitive user interfaces to ensure accessibility for non-specialist users.
Co-led project execution, system validation, and technical documentation.
Delivered a climate-adaptive, user-friendly simulation tool to support the efficient design of solar water heating systems across varied Iranian environments.
This project focused on developing a comprehensive software tool for the design and simulation of photovoltaic (PV) systems, aligned with Iran’s National Code 667. The tool was customized to incorporate regional climatic data and key design parameters, providing engineers with a reliable, standards-compliant platform for planning and optimizing PV installations.
Designed and implemented a user-friendly software platform for PV system analysis and sizing.
Ensured full compliance with Iran’s National Code 667 in all design and simulation workflows.
Integrated regional climatic, environmental, and geographic data to enhance simulation accuracy.
Led system architecture design and performed model validation.
Co-led project coordination, technical development, and final delivery.