Unlocking the secrets of vortex dynamics in stirred tanks through advanced modeling techniques.
Predicting vortex dynamics in stirred tanks presents significant challenges due to the complex fluid flow behaviors involved. The turbulence and chaotic nature of fluid movement in these systems make it difficult to accurately model and predict vortex formation.
Traditional computational fluid dynamics (CFD) simulations, while accurate, are computationally intensive and time-consuming. This poses a problem for industries that require quick turnaround times for simulations and optimizations.
Reduced Order Models (ROMs) offer a promising solution to the challenges posed by traditional CFD simulations. ROMs reduce the computational complexity by capturing the essential dynamics of a system with significantly fewer degrees of freedom.
ROMs are created by identifying the most important modes or features of the system and constructing a simplified model that retains the critical characteristics of the original system. This approach not only speeds up the simulation process but also makes it more feasible to conduct multiple iterations and optimizations.
At Ozen Engineering, Inc., we leverage our expertise in ANSYS simulation software to develop precise and efficient Reduced Order Models for vortex prediction in stirred tanks. Our process begins with detailed CFD simulations to understand the fundamental behaviors of the fluid system.
We then employ advanced model reduction techniques to extract the key dynamics and construct a ROM that can predict vortex formation with high accuracy. This ROM is validated against experimental data to ensure its reliability and robustness.
In this application we use the same stirring tank example that we demonstrated how to create a Fluent model for vortex prediction1. Tank is unbaffled cylinder, agitated with a Rushton turbine impeller. The geometry, and the mesh details on the cross-sectional plane are shown in Figure 1
Figure 1. The Geometry Model and Mesh
The Ansys Workbench project layout is shown in Figure 2. The geometry is connected to the Fluent session where the necessary model settings and input/output parameters are described1. The 3D ROM module is then dragged and dropped to the project screen, by which the 3D ROM module makes an automatic connection with the rest of the project.
Figure 3. The Ansys Workbench Project with the Addition of 3D ROM
Before doing the Design of Experiments (DOE), a better practice is to solve the first data set based on the inputs (DP0). This allows to monitor the simulation, post process, and determine if there are issues. Once this simulation complete, the parameters screen shows the corresponding output parameters (Figure 4).
Figure 4. The Input and Outputs for the First Simulation (DP0).
The properties of the DOE can be adjusted with the table (Figure 5) which appears once the Design of Experiments (3D ROM) is clicked. The default selection of "Optimal Space-Filling Design was used with 32 number of samples. This number is calculated by the software based on the number of input parameters. In this case, we have 4 inputs, and 8 configurations per input makes the total number of samples.
Figure 5. The DOE Settings
Once, the DOE simulations complete, green check mark is visible as in Figure 3. The designs with the associated results are available to review (Figure 6)
Figure 6. The DOE Results Screen
The next step is, click the ROM Builder, and Export ROM (Figure 7). There are two options to export; either the roms file or the fmu file. The roms can be read with a new Fluent session. The fmu can be imported to the TwinBuilder for further analysis. In this example we exported in roms file.
Figure 7. Export ROM
The final step is, open a new Workbench file, bring the Fluent session, and import the ROM. Once the Fluent session opens, click the Reduced order Model in the Models menu. In the panel click the "Evaluate" tab. Chose the suitable post processing tool, such as contours and select the ROM Cell Functions to be visualized. Changing the parameters used for ROM creation, the effect of that can be observed. Figure 8 shows that, lower speed does not generate a vortex, whereas high speed does.
Figure 8. The Evaluation of ROM for two different cases
From this point on, user can explore many different combinations of the input parameters. The result is available within seconds on the screen.
The primary benefit of using ROMs in vortex prediction is the significant reduction in computational time and resources. This allows for rapid prototyping and optimization, leading to shorter development cycles and cost savings.
Additionally, ROMs enable real-time simulations and control, which is particularly beneficial in industrial applications where quick decision-making is crucial. The ability to conduct multiple what-if scenarios also enhances the overall design and operational efficiency.
The details of the of the ROM generation can be found in the below video.
The vide can be accessed from Ozen Engineering YouTube channel with the below link:
Creating a Reduced Order Model for Vortex Prediction in Stirred Tank
Reference:
Video: Simulation of Steady Vortex in a Stirred Tank
We offer support, mentoring, and consulting services to enhance the performance and reliability of your hydraulic systems. Trust our proven track record to accelerate projects, optimize performance, and deliver high-quality, cost-effective results for both new and existing water control systems. For more information, please visit https://ozeninc.com.