[davidvalladares]

Hi, I´m brand new in CFD and I´m trying to simulate a pump impeller. I did all de tutorials I found and so far everithing was ok.

Then I modeled my oun impeller (the impeller exist and works fine, I work for a pump factory and this is the impeller of one of our pumps) but when I´m trying to simulate I found that the flowlines crosses the stationary walls.

What am I doing wrong ???

I attached a picture of the disaster

Impeller diameter: 178mm
Q = 125 GPM = 0.0079 m3/s
H = 60 mca
n = 3550 rpm = 371.75 rad/seg
static pressure at outlet lid = 637414 Pa
Internal analisys
Global rotating
Resolution Level= 4
Minimum gap size= 0.04m
Minimum wall thickness= 0.01m

[Boris_M]

Hi David,

this should not happen. At least for me it looks like it is outide of the model completely but you use internal analysis. I don't know if there is any other geometry you are hiding so can you check the geometry using the "check geometry" tool and show the fluid volume?
That should show if there is actually a fluid volume in the region of the flow trajectories.
In post processing the graphical modelling of the flow trajectories can in very rare cases jup from the inside to the outside of the model but that only happens if it is an external simulation and it is just a graphical issue. It can happen that when the trajectories are created and the software calculates the trajectories that it jumps from one cell to the other when creating them but accidently through a thin wall but then there is maybe 1 or 2 trajectories outside the model. In your case it looks more like all of them are outside.

I would assume there is some geometry problem with your model, either geometry that is deactivated in the component controls and therefore not limiting the flow region or for some reason ignored by the mesher due to import problems if it is an imported geometry or it is a surface model either due to the importing too or by design.

Boris