Sounds like that's how it happens due to conditions present in the system.

"Random coalescence" may be a problem source because if it s a dense plume, random coalescence will occur quite frequently..

well, the thing is what would be the difference from simply set d0=5 mm then?

I mean the thing simply goes to d==dmax almost automatically as the alpha air progresses...

for the values I leaved the default ones but even in the tutorial it completly goes to dmax automatically.

this is only after 3 seconds of simulation (tutorial) https://cdn.discordapp.com/attachments/726377729229979669/1369683032130191451/image.png?ex=681cc03a&is=681b6eba&hm=383a3ce6f9ab440f029f4ade51dd8f7407296ffb03f3ca0c17ed86ed7a9b005f&

In the above figure, is alpha.air 1 in the whole domain? What's the value in the alpha.air 1 region? (Like U.air will be non zero even when alpha.air is zero)

Looking at the expressions for wake and random coalescence I must say these don't seem extremely well-founded on theory.. more like proportionality based models. So I wont be surprised if careful calibration is needed for coefficients. ( I can't find the cited paper immediately, to see how the original configuration was set up and benchmarked) https://cdn.discordapp.com/attachments/726377729229979669/1369690765130727575/rn_image_picker_lib_temp_9def6cb7-853b-4fd9-8a1d-c972cbc48f99.jpg?ex=681cc76e&is=681b75ee&hm=df9aecb46f2df765b879b21da0e15eecfef5f7fe2ba38759378e9cc002f085fa&

Seeing another issue with the defaults in the tutorial - the mechanism is turbulent breakup (Ut comes from sqrt( TKE)), but the sim is laminar... So no breakup to balance coalescence.

why the thumbs down :o?

Couldn't access, encountered lots of Korean

From a related experimental benchmarking paper of this work... https://cdn.discordapp.com/attachments/726377729229979669/1369695291783250062/rn_image_picker_lib_temp_3129cbe0-4d37-4ea9-94cc-509581644ad4.jpg?ex=681ccba5&is=681b7a25&hm=399d03efd49ce900dc3f85070c66846c853120bcb3046c1a8919fea466abaf52&

RC is 10x

my bad, the uni had acces and didnt realize it was not simply open, the page is in korean but the pdf on it in english

For random collision the default value seems 10x the specified one in paper https://cdn.discordapp.com/attachments/726377729229979669/1369695948275716147/rn_image_picker_lib_temp_3d4d1daf-4306-488e-98d1-db3049a39541.jpg?ex=681ccc41&is=681b7ac1&hm=1ece4a397992acf7e04520d4b91656994bf908d335a82d890589ac0ca04aeb1c&

running

same XD https://cdn.discordapp.com/attachments/726377729229979669/1369696751354773616/image.png?ex=681ccd01&is=681b7b81&hm=c354dbdeeec3310371c7e917e44d6d27014927a581e8e2bb610962b82a66f591&

I see

I will change turb also to see if it imprves

well..... something at least not completly dmax now... https://cdn.discordapp.com/attachments/726377729229979669/1369699202782855310/image.png?ex=681ccf49&is=681b7dc9&hm=51e2093139a6c7ecab1424308a599aa36ef8a57aa8136e3df47c193a53d7cf01&

Does it keep increasing even if you set ridiculously large values?

For dmax

nop!

it actaully does not go as low nor as high as the boundaries

There are a bunch of papers by these guys - along the lines of benchmarking the IATE model for different geometries etc. see if you spot a better suited set of coefficients for your config. (Like bubbly vs churn-turbulent), shape of column..

but at least I will add the things you spoted into an issue as they are clearly errors (y)

I guess it maybe better to just move the tutorial to RAS and change the coefficient