Dear colleagues
I am trying to simulate a multiphase heterogeneous reaction in fluent. in this reaction there is a mixture of gas above a liquid pool.
at first gas mixture is pure A and in presence of liq B will be produced.
simulation keeps going truly but after a while B appears in liq!
I didn't activate interphase mass transfer.
why this is happening?
plz take a look at attachement.
Hi Mahdi, I can't download the video. But do you have a reactio defined in your model? If yes, think about: in a Volume of control that contains the interface, you will have A in thepure gas ,and the liq. Exactly at the interface you have A, and the liq in contact, hence, at this point B can be generated without a mass transfer. As we have a eulerian framework, the volume of control have an average of gas and liq and your source term that generates B. In this Volume you have all condition needed to generate B . It is a theory.
Dear Ricardo
many thanks for your fruitful helps.
but I have encountered another problem.
I have attached two pictures below.the inter phase is stagnant and in middle of tank.
up to a certain time we have a gas product species profile is ordinary and in order of 10^-7.
but in a spur of the moment dots of that species appears with an order of 10^-1(0.25 to 0.5) in liquid where there is no inter phase.
please take a look at this two pics.
regards
Oh , I've forget of this detail. The code as Fluent and CFX have by default a cliping value to volume fraction to avoid roud-off errors. Your source must to has a limiter, a IF or MAX or MIN statement to deal with this values. But is very strange these two points in red! How is your mesh? In these localtions the elemnts are OK? there are some distortion or acute angles?
Dear Ricardo
nice answer!
meshing is structured but it was coarse!
I refine it was better but that odd error happens again.
changing to VOF obviate this error, It looks like that eulerian approach assumes it as particulate!
regards
Oh it sounds good, but that error must does not happen. But I have some bad experiences with ANSYS codes, maybe a curse :D. Sometimes, with a very specific setup and a CORRECT mesh. Bad results were obteined, soft craches, or abnormal termination. Answer : BUGS!!!
Changing one parameter, all followed OK. In release 10 of CFX the Rhie-Chow P-V coupling had a error in coefficients of a density. I was work with iron-making,a 7000/1 density ratio!!! My convergence was slow. In the new services packs and releases it was fixed. As Fluent was acquired by ANSYS, and the releases are delivery very fast - we have the new R16 - it is possible thta some like that it has happened.
But , if it is OK now, it sounds good. Good look!!!
- Badges
- Science topic