CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   Laminar unsteady flow convergence (http://www.cfd-online.com/Forums/fluent/64595-laminar-unsteady-flow-convergence.html)

jpinho May 15, 2009 12:00

Laminar unsteady flow convergence
 
Hi there,

Im doing a simulation for a heat exchanger where i heat water inside an aluminum block. The flow rate is low i assume laminar model.

When I try to solve it steady, the solutions converge and i get fine results. but when i tried to do it unsteady they dont.

I started with a condition where there was no heat coming to the water, and then i assumed a heat generation of 20000 w/m2 going towards the water with 20 time steps of 3 seconds interval solving on maximum 40 iterations each.

In the beginning the water starts heating but after 6 time steps it doesn't converge... it prompts error on minimum temperature 1k...

any idea?

thanks
jorge

ak6g08 May 15, 2009 14:43

hi jorge,

minimum temperature 1kelvin I think means that somewhere in your solution, you are effectively getting values of temperature under absolute zero, which is physically impossible and thus fluent doesnt like it.

any error that fluent pops up is an error from your grid generation/boundary definitions or boundary conditions in my experience...try changing the boundary conditions around to pinpoint the error, i.e. make all your walls fixed temperature boundary conditions, and then one by one make them fixed flux (at zero) boundary conditions to see when the error comes up and that will tell you if its a boundary issue which i suspect it is. otherwise, try decreasing your timestep to see if that makes a difference.. hope that helps

akour

jpinho May 19, 2009 10:16

Thanks Akour,

I found the problem.
The mesh quality was really poor and there was a discontinuity in some points of the geometry.. i will start it over..

thanks again
jorge


All times are GMT -4. The time now is 00:46.