Learning Forum

[Jack5864]

Hi there,nnI'm running a fluent simulation where the transport equations for my 2 UDS variables are highly coupled. The UDS diffusivity for the first is proportional to the value of the second, and the mass flux term for the second has a term proportional to the gradient of the first. This results in a highly coupled system of differential equations, which will not solve at the moment. Instead, the solution very quickly diverges.I was wondering if anyone had any advice on how to go about improving the stability of the solution.Thanks,nJackn

[Surya Deb]

Hello, nAre you using a Coupled PT solver? nDo you solve the UDS and the flow variables simultaneously? If the UDS properties are not tightly coupled to the flow, then can you converge the flow field first, then freeze the flow and solve for the UDS?nWhat discretization schemes are you using?.Regards,nSuryan

[Jack5864]

Hi Surya,nnThanks for your message. The flow profile is relatively simple, so I use a SIMPLE solver for the pressure-velocity coupling. I also do as you mentioned, solving the flow first and then the UDS equations. The problem is that the UDS equations are tightly coupled together I believe. nThe discretization schemes I use are Second-order Upwind for my UDS variablesnnThanks,nJackn

[Rob]

Have you coded in stability terms for the two UDS equations? You might need to relax the relationship to improve the stability. n

[Jack5864]

I haven't coded that no, how would I go about doing that?n

[Rob]

Carefully! Simply you need to reduce how much you change each value each time in much the same way we do in the main solver: how is far more complicated. n

[YasserSelima]

You basically need to run a solver inside Fluent solver!! Good luck with thatnGoogle github CFD C++ code ... you might find a CFD code with 2D solver and you use it as starting point. n

[Jack5864]

Thank you both,nnI'll do some digging to see what I can find!n

[Zhangmin004]

Hi there,I'm running a fluent simulation where the transport equations for my 2 UDS variables are highly coupled. The UDS diffusivity for the first is proportional to the value of the second, and the mass flux term for the second has a term proportional to the gradient of the first. This results in a highly coupled system of differential equations, which will not solve at the moment. Instead, the solution very quickly diverges. I was wondering if anyone had any advice on how to go about improving the stability of the solution. Thanks,Jackhttps://forum.ansys.com/discussion/24336/highly-coupled-uds-transport-equations

Dear Jack,nI just encounter the same problem as yours. Have you solved it?nI do not find out the coupling solution method for coupled UDSs in Fluent. I will really appreciate if you could tell me any advice about this problem?.Best regards,nMinn