CFD analysis of heat transfer enhancement in plate heat exchanger

Plate heat exchangers are key components in various industrial processes, offering efficient heat exchange between two fluids while maintaining a compact and space-saving design. These devices find extensive use in industries such as HVAC, heat pumps, chemical, food processing, and power generation, where temperature regulation and heat transfer are critical. They are considered due to high thermal efficiency, low maintenance requirements, and the ability to handle different flow rates and temperature differentials. Characterized by their stacked plate configuration, these exchangers facilitate efficient heat transfer through a large surface area, enabling rapid exchange of thermal energy between fluids without direct contact. However, in the quest for enhanced heat transfer efficiency, continuous research is going on to explore innovative ways to further optimize their performance. This is where nanofluids come into play. By introducing nanofluids into plate heat exchangers, we aim to boost heat transfer rates, which can lead to improved energy efficiency, reduced operational costs, and enhanced overall process effectiveness. Therefore, delving into the Computational Fluid Dynamics (CFD) analysis of plate heat exchangers with low concentration alumina nanofluids offers a promising path to understand and estimate potential benefits of nanofluids in heat exchange enhancement and energy efficiency improvement.

Modeling and CFD Simulations

For this purpose, a plate heat exchanger model is developed with FreeCAD software and OpenFOAM is used a solver. After considering literature data and plate heat exchanger drawings available online from heat exchanger manufacturers like Alfa Laval, the dimensions of the plate were decided. The design was simple, consisting of a single chevron plate with inlet and outlet ports. Gaskets were also added to replicate the real life plate design. The setup involved hot and cold fluids flowing in opposite directions on each side of the plate. In the figure below, the plate geometry is shown along with the flow pattern.

After modelling the plate, it was imported to the solver to add fluid domains, set the boundary conditions, generate high quality mesh and conduct the transient three-dimensional CFD simulations utilizing the implicit time stepping method. Some assumptions considered for simulation are as shown below:

1. Loss to surroundings is neglected
2. Flow enters normal to the boundary faces
3. Turbulence intensity at inlet is set at 5%

The computational mesh was generated employing high-quality polyhedral cells while accurately capturing the boundary layer and reducing the computational time. The boundary conditions set for the simulations are shown below:

1. Velocity is input parameter for at Inlet for both hot and cold side and pressure is provided as outlet parameter.
2. All other walls are insulated, the heat transfer takes place across the chevron plate only.
3. Counter current flow set-up, where the hot side fluid flows downwards and cold side fluid moves upwards.

Demineralized water is considered as the hot side fluid and the cold side fluid is a mixture of Water:Ethylene glycol (50:50 v/v). The low concentration nanofluids are also formulated with the same base fluid water:ethylene glycol (50:50 v/v) with addition of alumina nanoparticles (~1% wt, 100 nm diameter) and is used as cooling liquid. The heat transfer performance and thermal maps for both base fluid and nanofluid are compared to understand the potential improvement in heat exchange performance when nanofluids are used as cooling liquid. The input parameters used to perform the simulations are shown in table below:

The experimentally measured demi-water, base fluid and nanofluid parameters that were used to perform the simulations are shown in table below at their respective inlet temperatures:

Since the alumina concentration is quite low, there is no significant change seen in the thermo-physical properties except for thermal conductivity.

Results and conclusions

Using these values and inputs, the modeling of turbulent flow was accomplished through the Shear Stress Transport (SST) formulation for the k-ω turbulence model. Although flow is not turbulent, turbulence equations (k-ω SST) have been used in order to input the surface roughness parameter. The results after convergence are shown below as thermal maps:

Figure depicts the temperature distribution on the hot fluid side of the plate. The figure clearly shows that nanofluids reduce the temperature of the hot fluid (demi-water) more than the base fluid case.

Figure depicts the temperature distribution on the cold fluid side of the plate. The figure clearly shows that nanofluid increases the temperature of the cold side more than the base fluid case, thus transferring more heat and showing improved overall heat transfer coefficient.

Calculating overall heat transfer coefficient (U) based on the steady state simulations, results in higher values for nanofluids compared to base fluid on the cold fluid side at similar pump powers and Reynolds numbers as shown in figure below:

With low concentration alumina nanofluids, a higher heat transfer coefficient and energy efficiency can be achieved. More experiments are being carried out at Synano labs to test plate heat exchangers with higher nanofluid concentrations and surface modifications of the plates to further improve the heat transfer efficiency.