Discussion Forum

This particular subject isn't in my specialty, but... considering how very different (at least, subjectively) that your results appear, I would think that if either one of these two models is even approximately correct, then it should be pretty easy for you (as a scientist, engineer, and/or mathematician, regardless of your level of experience) to spot, with some confidence, which one of these must be wrong (while the other one merely may be wrong). If so, and if I were you, I would then look more closely at the one that I was sure was wrong, to try to find an error in the problem specification. Once you find that error, fix it and re-run that model and compare it again to the one that wasn't so obviously wrong. Eventually, you just might get your two models to agree.
Added: If you don't have access to the ANSYS model (and software) to work with yourself, then I suppose you can't study it in detail, so I guess you'll have to decide (based on the research you mention) whether you believe it or not. If it really matters to you, then you might want to contact the authors of the research and see if they are willing to explain to you exactly how they configured their model. Then you can compare it to your own, one feature at a time, step by step. One last thought: I don't know much about mixing, but your Comsol model looks like some substantial mixing is occurring. In the other model, not so much. I'm guessing this is the key to deciding which one must be wrong. Finally: If you decide your Comsol model must be wrong but if you work and work and you just can't find the error, then consider posting the .mph file to the forum, so others here can take a look at it. Hmm, I just found this on the internet. Is this the research you mentioned? **https://www.mdpi.com/2072-666X/13/9/1375/pdf **

Yes that is the research. Also this is the link of .mph file

https://www.dropbox.com/home?preview=Research.mph