[ALBATTROSS]

Hello all,

I'm solving a transient natural convection problem using finite-volume method and comparing it with another results with different approach. The problem is in order to recover their results I should do a lot of iterations (2000 ) in each time-step in FLUENT's fully implicit transient solver, otherwise, the transient solution comparison has a very large error (for example using 20 iterations instead of 2000 is very off) but the steady-state results have very small error. The problem is, in order to be able to iterate that much I should set equations residuals criteria as low as 10^-12. My conclusion is this problem is very stiff so the equation residuals are as low as 10^-12 but the variable residuals are 10^-3 I want to justify this idea so I need to prove why the problem is stiff, can you guys help me finding sources that justified a problem like this. or any past experience with situations like this that can ignite another idea?

Thank you