Particle Trace Losing Continuity

Using SimScale Fluid Flow / CFD
1

I am designing an axial jet engine and the simulation I just ran using this as a guide (How to Run a CFD Simulation of a Centrifugal Pump - YouTube) is not showing continuity through the rotor of my jet. In other words the particle trace is terminating in the rotor and is not continuing to the outlet. I am worried about trusting the rest of the data off of the simulation because of this. I would love to know what is going on here. Thank you.

Here is a link to the project https://www.simscale.com/workbench/?pid=5137867480164366136&mi=spec%3A0dd0b2eb-eee9-44ff-b29e-fabd30de649f%2Cservice%3ASIMULATION%2Cstrategy%3A1

2

Hi, and thanks for using the forum!

I agree that this should not be the case. I will report this as a possible bug and come back to you when I have news from the dev team.

1 Like
3

Thank you, excited to see what you can find. I am worried about trusting my pressure and velocity data so I am curious if you all think that that data can still be trusted. Right now it looks like cavitation is occurring at the trailing edge of the rotor tip, which based on my numbers shouldn’t be the case but…

4

I think that the problem is in the particle trace from the postprocessor.

As for the trust in your numerical results for the flow fields, you should perform activities such as:

  • Check residuals
  • Measure convergence with result control items
  • Perform mesh independence analysis
  • Perform sensitivity analysis against the input variables
5

I see, that is a little reassuring.

It looks to me like everything is converging though I am not a data scientist. Except for the domain and outlet. Would you mind taking a quick look at those and seeing is they appear correct to you?

Are residuals supposed to show a convergence as well? They seem too.

I am not sure how to perform a mesh independence analysis or a sensitivity analysis against the input variables. Has simscale released and videos or articles on how to do this? My mesh was divided into 3 different pieces because of the rotating zone. I found this a little odd.

Thank you.

6

Hi IronManDylan,
This convergence document I found helpful and to get a feel of what was reasonable and what to expect for the convergence plots I also referred to the convergence plots from a Simscale tutorial that was similar to the project.

Kind regards
Ted