payday
Mechanical
- Feb 3, 2015
- 1
So in the last few weeks, I have basically taught myself ANSYS, with the help of Google and the ANSYS manual and tutorials. These questions are going to focus mainly on the CFD portion. Some are general, and some are super specific to what our company is trying to do, so anything you can help with is much appreciated.
1. What are some good rules-of-thumb/practices for evaluating your mesh before using it in a run? When do you know that it's not good enough and you should try something else?
2. How do you choose the initial mesh sizing?
3. For mesh convergence studies, it seems like it would take a lot of extra time to change the mesh a little, run it all over again, and repeat it several times. Is that the best way?
4. When doing mesh convergence, how do you select mesh sizes to test.
5. I'm using this mainly for Turbomachinery applications. When modeling an expander, it is rotating with the flow, but removing energy. How is this modeled?
6. I know you are supposed to de-feature your part to simplify the flowpath, but are there any features (fillets, etc.) that are important not to omit?
7. What is a good rule of thumb for selecting a timestep?
8. What does a timestep mean for a steady state model, physically?
9. For one model, I am attempting to use materials from an imported Peng-Robinson dry hydrocarbon library. They are modeled as real gases. In modeling a compressor or expander, I continuously get the error "Bounds Limit. Density Derivative WRT Pressure at Constant Temperature." What might be causing this?
10. What is the best way to model turbomachinery: rotating reference frame, deforming/moving mesh, immersed solid? Or at least, what are some pro's and con's of these methods?
11. What are some good ways to improve convergence. Often with the above Peng-Robinson materials, when it doesn't break due to that density error, it converges to about 10^-3 but then oscillates and doesn't go anywhere.
12. What are some of the most useful/common Post-processing tools?
I know this is long, but, like I said, any help is much appreciated.