[ajithnair]

Will adding relaxationFactor in fvSolution affect the accuracy of solution? If so what is a safe value of relaxation factor.



I am doing a simulation on methane combustion in a Archemedian spiral channel and by changing relaxationFactors the result keeps changing. The flame stabilizes at different locations just by changing the relaxationFactor. Results are only obtained for relaxationFactor < 0.7. Any value above it causes the chemical time step to increase to an order on e-11 and solution diverges.

-------------------------------------------------------------------------------------

relaxationFactors
{
equations
{
".*" 0.5;
}
}
--------------------------------------------------------------------------------------

[ajithnair]

I found some material about relaxation factor here



https://www.simscale.com/docs/simula...ation-factors/

Originally Posted by ajithnair

Will adding relaxationFactor in fvSolution affect the accuracy of solution? If so what is a safe value of relaxation factor.



I am doing a simulation on methane combustion in a Archemedian spiral channel and by changing relaxationFactors the result keeps changing. The flame stabilizes at different locations just by changing the relaxationFactor. Results are only obtained for relaxationFactor < 0.7. Any value above it causes the chemical time step to increase to an order on e-11 and solution diverges.

-------------------------------------------------------------------------------------

relaxationFactors
{
equations
{
".*" 0.5;
}
}
--------------------------------------------------------------------------------------

[piu58]

Relaxation should not influence a steady state result. If your case is straightforward and relaxation is not necessary: Introducing it increases calculation time, nothing else.

In transient simulations under-relaxation acts as some kind of slow motion and may be smear effects.