CFD Online Discussion Forums

CFD Online Discussion Forums (
-   Main CFD Forum (
-   -   Pathlines Problem (

Robbb May 7, 2013 04:25

Pathlines Problem
Hello everyone. I'm simulating a fluid flow through an helicoidal heat exchanger. The fluid (liquid) gets heated along the exchanger. When I'm post-processing the simulation and I'm trying to display pathlines starting from the inlet to calculate total temperature profiles along the exchanger, everything is ok but there's always one single pathline which doesn't escape the exchanger, or better.. this single pathline takes much longer to escape the exchanger. I've tried to change the flow rate entering the exchanger, but there's always the same problem. How can I fix this?

triple_r May 7, 2013 10:20

Unfortunately, if there are recirculation zones in the solution domain, this can happen a lot. If the solution is steady state, then the recirculation zone should be a captured eddy (no mass exchange with outside of eddy), but because of numerical errors, even in steady state the eddy can exchange mass with outside. When this happens, and by bad luck, the particle that you are tracing enters the eddy, it takes a long time for it to get transferred outside (or it gets captured inside, recirculating over and over).

You can try refining your mesh near the recirculation zone, to remove some of those numerical errors, or let your solution run longer to get a better steady-state convergence. The other way is to trace a lot of particles, and ignore the ones that have a much longer residence time.

I hope this helps

Robbb May 9, 2013 08:49

I tried to extend the convergence to from E-06 to E-8, but it still doesn't work. I should probably modify the mesh, but the geometry is really complex and I actually can't understand where are the problems. I think I'll try something else.

All times are GMT -4. The time now is 11:00.