CFD Analysis of Cooling in Adiabatic Expansion

Hello everyone, this is Mohsen Seraj. I'm a senior application engineer for ANSYS, working with Ozen Engineering. Today, I want to talk about CFD analysis of cooling in adiabatic expansion. Consider a system with a flow coming through a narrow way.

When it enters an opening, if the system is closed (isolated) and there is no exchange regarding heat with the surrounding, we should see that the pressure and temperature drop. This cooling happens without any additional energy to the system.

We can use this pressure and temperature drop for cooling in that situation and even for previous processes in industrial applications. This also happens in geology and nature. For example, when air moves up, warm air is cooled, and the temperature drops.

We should see more water droplet formation as it goes up and most likely a cloud phase as it goes down. We can use this method in data centers, HVAC, air conditioning for buildings, and cooler and condensation through the adiabatic cooling system.

This method is very energy efficient, as we don't need any additional work on energy for the system. It could be a closed, isolated system, and the operation condition is very low. This has a significant impact on water usage and the environment.

This method can save billions of liters of water annually for cooling in the cooling system for data centers or other industrial applications. Now let's discuss the difference between adiabatic cooling and the Joule-Thompson effect.

Although they are similar thermodynamic processes that reduce gas temperature, they work differently. Adiabatic cooling mostly relies on the expansion of gas in a closed system without heat exchange with the environment.

The Joule-Thompson effect mostly relies on the sudden change in temperature and pressure, not necessarily with a significant volume change. Let's see how we can model this effect in ANSYS Fluent.

I have prepared a simple problem with a smaller part (50 mm radius) and a larger size part (200 mm radius). Each part is long enough for fully developed flow. I assigned names to the model: inlet, outlet, and walls. In ANSYS Fluent, I use a steady-state pressure-based solver.

I read the mesh, and the mesh quality looks good, with an orthogonality above 0.4, which is very good. The minimum orthogonality is also acceptable. The mesh info shows approximately 60,000 polyhedral cells. For material modeling, I use the ideal gas model.

I consider the second order of wind condition. For report definition and monitors, I defined max temperature, mean temperature, and max velocity in the domain. I initialize the model with gauge pressure, zero velocity everywhere, and a total temperature of approximately 80 degrees Celsius.

I define a plane symmetry (YZ plane) and set up rigid walls with temperature, pressure, and velocity contours. Now, let's run the simulation for 500 iterations and check the rigid walls, mass flux imbalance, and heat transfer rate.

The mean temperature is approximately 11.8 degrees Celsius, which is very good. The max temperature is 80 degrees Celsius, and the velocity is about 370 meters per second. In conclusion, when we consider ideal gas, we can capture minimum temperature much lower than the initial temperature.

This adiabatic expansion causes not only temperature drops but also pressure drops. This technique can be used in industry for cooling and temperature drop without any additional energy system, making it cost-effective.

I hope this example demonstrates the proof of concept of adiabatic cooling due to adiabatic expansion. Thank you for following up with this example.