Technical/Scientific Challenge
In this project, EYS Makina is challenged to respond to a customer request in a very short amount of time. The mixing characteristics of large-scale renewable biogas energy plant reactors have to be analyzed before the construction and deployment of mixers. Our R&D team rose to this challenge using open source CFD software and HPC resources, analyzed multiple scenarios within a short time, and continues working on multiple different scenarios for our future needs.
Customer requirements. Short response time. Large scale flow physics problem. Solved with the help of open source CFD software and HPC resources.
Solution
Analyses are carried out in different configurations using dive and tangential angle parameters. The simulations have to be carried out in three dimensions and unsteady using an accurate turbulence model. In this study, the large eddy simulation (LES) approach is followed for studying three-dimensional turbulent flow. The most optimal results were achieved when the number of cells per processor was around 10,000. It is also noted that the number of cores per node is important, as it is appropriate to work with a full node from the cluster to achieve optimal communication load. The flow is modeled as periodic because it contains periodic repetitions.
Figure 1:Mixer positioning.

Figure 2: The full model of the digester.

Figure 3: Snapshot of the flow field at different cross sections.

Figure 4:Instantaneous flow field represented with velocity magnitude at constant z-section.

Business Impact
EYS Makina benefits from HPC resources in both design and application cases. In this project, EYS used HPC resources to investigate and optimize mixing processes inside biogas digesters. Parallel, scalable, open-source CFD tools are utilized to achieve this goal.
While tackling this practical engineering problem, EYS Makina is also aiming to improve their HPC skills and reach a higher and more efficient level of HPC resource usage. In order to do so, EYS Makina:
- closely monitored scalability and parallel performance metrics.
- try to increase utilization of automated operations, such as on-the-fly data analysis, file transfer processes, or using certain software applications remotely on HPC servers, etc.
- try to streamline steps in their flow chart, aiming to minimize human intervention in the process (e.g., geometry generation, mesh generation, solver setup, etc.).
Benefits
- Reduced response time to our customers
- Increase in customer satisfaction
- Strengthening market competition
- New customer demands on digester analysis
Keywords
Contact:
Engin Orçun Kozaka, EYS, okozaka@e-y-s.com
Ayse Gul Gungor, ITU UHeM, ayse.gungor@itu.edu.tr
