[raskolnikov]

Hi all,



i have a problem with a flow problem which i will try to explain below (The number of volume mesh elements is 30 Mil.) :



it is a compressible flow, medium is air. For the simulation i chose the ideal gas formulation. The temperatures at inlet goes over 1000 Celcius, and the velocities are also high but they are still under Mach Number=1 over the time (t_end=10 s, Time Step= 1e-4s)



My initial temperature for the fluid phase, also the solid parts is 25 Celcius.

The problem occurs at both flow formulations in different ways: segregated vs. coupled.



In the segregated case it could be seen that mass conservation is good enough. But lower temperatures (under 25 Celcius - minimum 5C) occur in the domain, which are unphysical.



With the coupled solver the minimum temperature in the domain is 25C, which is realistic. On the other hand a big mass imbalance could be obtained, almost the half of mass flow is lost. It can be also seen at the outlets with really low mass flow values with the coupled flow solver.



I know coupled solver is more stable, in this cas for the energy equation, but what about the mass balance? How can I improve the mass balance?



Or do you any suggestions to solve it, maybe some clues to reach a stable solution?



Thanks in advance.

[f-w]

Have you tried initializing the transient case with a converged steady-state solution?

[Maddin]

Without more information I would say 25°C initial but 1000°C in simulation you may shoudl consider different initial, more steps for ramping - CFL/relaxing factors you could increase.

[Fluidalessio]

Hi, I have a similar problem and I'm using a supercritical fluid.. I know it's been a while but how did you solve it?