- #1
kdgarg
- 1
- 0
How can we predict the actual test performance of pump from CFD. I need some formulation/calculation method to predict closure.
Centrifugal pump's Test performance prediction from CFD
- Thread starter kdgarg
- Start date
Click For Summary
SUMMARY
This discussion focuses on predicting the actual test performance of centrifugal pumps using Computational Fluid Dynamics (CFD). The recommended approach begins with a meanline or 1D code, which utilizes a "velocity-triangle" method for initial calculations. Following this, a model is created and meshed using ANSYS ICEM to achieve optimal hex mesh quality, crucial for accurate CFD results. The final step involves solving the model with moving boundaries applied to the rotor, yielding results that closely align with experimental data.
PREREQUISITES- Understanding of Computational Fluid Dynamics (CFD)
- Familiarity with meanline and 1D analysis techniques
- Proficiency in ANSYS ICEM for mesh generation
- Knowledge of rotor dynamics and moving boundary conditions
- Research advanced techniques in Computational Fluid Dynamics (CFD) for pump performance prediction
- Learn about meanline and 1D analysis methods for centrifugal pumps
- Explore best practices for mesh quality optimization in ANSYS ICEM
- Investigate the application of moving boundary conditions in rotor simulations
Engineers, CFD analysts, and pump design specialists looking to enhance their understanding of performance prediction for centrifugal pumps using advanced simulation techniques.
- #2
minger
Science Advisor
- 1,494
- 2
There are lots of tools developed for just this. Typically our approach for compressors is to start with a meanline or 1D code first. Basically it's a glorified "velocity-triangle" approach to solving, but will get you very close.
From there a model is created and meshed. For the best mesh quality, we bring it into ANSYS ICEM in order to get a full hex mesh. Mesh quality if very important with CFD; always.
At that point it's brought into your solver and solved. Moving boundaries can be applied to the rotor and from what I've seen, results can be quite close to experimental values.
From there a model is created and meshed. For the best mesh quality, we bring it into ANSYS ICEM in order to get a full hex mesh. Mesh quality if very important with CFD; always.
At that point it's brought into your solver and solved. Moving boundaries can be applied to the rotor and from what I've seen, results can be quite close to experimental values.
- #3
FredGarvin
Science Advisor
- 5,097
- 10
minger said:
And then we put it in the test cell and tell Minger that he's out to lunch
Similar threads
- · Replies 8 ·
- · Replies 2 ·
- · Replies 7 ·
- · Replies 4 ·
- · Replies 15 ·