|
[Sponsors] |
December 5, 2000, 09:45 |
Waterinjection
|
#1 |
Guest
Posts: n/a
|
I am working on a model using water injection to cool a gas flow through a pipe. Applying the lagarangian multiphase/droplet model I have defined the droplet properties and the heat of evaporation, boiling model is activated, and the droplets are evaporated into scalar H2O. However using the mass flow of water injection required to cool the flow, found using simplified equations, does not cool the flow as much as expected. Neither does the water injection seem to have much impact on the gas flow.
Does any one have any comments/experienced similar things. |
|
December 8, 2000, 05:19 |
Re: Waterinjection
|
#2 |
Guest
Posts: n/a
|
Are you using the momentum transfer from the droplets to the gas phase ? If this is not used, then the droplets cannot influence the gas flow. Additionally, if you are doing transient simulation, you should use small enough time step (droplet not passing through more than one cell lenght during one time step). Also the grid should be relatively fine in the area where there are droplets (e.g. 2mm grid size is relatively fine).
Regards |
|
December 8, 2000, 10:10 |
Re: Waterinjection
|
#3 |
Guest
Posts: n/a
|
Hi Ossi,
I have activated the momentum transfer, standard option. At the moment I am running the case as steady state, and I am working on a fairly coarse mesh (approx 10 mm cells). I have had some advice from CD on activating switch 199, to prevent condensation. However there seems to be some issues with regards to boiling model on my current Star installation v3.100. The mystery bit is, when the switch 199 has been activated on CD's v3.100B I can run the analysis and get more sensible results. I will no go on to refine my model and hopefully get a result. Thanks for the tips, Paal |
|
|
|