Modelling underwater sea current

shaman · May 23, 2016, 17:39

Hi all, i need to model thermal dilution of cold water discharge into seawater.
I have 3 question
1. How to model the seawater current which has a boundary layer like velocity profile... ie max at value A at surface and reduces as it reaches bottom. I have tried to model this as velocity inlet and outlet with same for domain initialization but the steady state analysis converge to a different solution. I need to maintain this current throughout the open sea domain. Can i use source.. if so how to maintain velocity profile.
2. My open sea domain is a hexagon 150m radius .. how to model translational periodics so my body of water represents d sea with underwater current mentioned above.
3. How to introduce scalar variable to track the behavior specific to the cold discharge water

ghorrocks · May 23, 2016, 20:49

1) You will only get your intended velocity when you set the simulation up such that it gives the profile - in other words you have an appropriate turbulence model, turbulence conditions and other variables. So if you don't get the right profile it is because you have not set the model up such that it gives you the profile.

2) Make the matching parallel faces translational periodic pairs. So you will have 3 pairs for your hexagon.

3) Simple way: use an additional variable with convection and diffusion. But this will not account for the variation in material properties due to the difference in composition. If you want to account for this (to model things like the density difference due to salinity content) then use a multicomponent mixture.

shaman · May 23, 2016, 20:55

Tqvm.. ghorrocks..
The sea current i m looking to model is less complicated as compared to prevailing sea current with the earth movement..

shaman · May 23, 2016, 20:57

I need a body of water moving at a velocity profile of my making vs depth.
The profile is from metocean study so it is reasonably accurate... and below 0.8ms-1.

shaman · May 23, 2016, 20:59

Isnt it possible to introduce a source term to that body of water... n the discharge water resolved as normal

ghorrocks · May 24, 2016, 00:57

Yes, you can apply a source term. But if you apply a source term everywhere it will probably interfere with the flow around the region of interest to you. It could be better to put the source term on the periodic face only to pull the velocity to the profile.

If you know the velocity profile then why not forget about the periodic faces and just use an inlet with a prescribed velocity profile? Then you can specify the the profile easily.

shaman · May 24, 2016, 01:46

Without source term wont the velocity at the discharge location diffuse due to large seabody. I could initialize the seabody with similar velocity profile. With no slip floor

ghorrocks · May 24, 2016, 01:50

If you have the correct turbulence model and conditions then the flow should not diffuse too much. This was my first point in the first post - you should try to make the physical model correct first, and then the need for a source term to enforce the desired velocity profile is greatly reduced.

shaman · May 24, 2016, 02:30

Only have metocean data as ref for velocity profile.. have to make do with source term .. can the source term be velocity with function of height....and initialize sea body similarly

ghorrocks · May 24, 2016, 07:26

I do not understand what "metocean data as ref for velocity profile" means.

Yes, a source term can be a function of height (or just about anything else for that matter).

shaman · May 24, 2016, 08:38

Metocean is d oceanographic study.. simulation conducted to simulate design criteria for specific return period ranging from 10... 50... 100 and in between.... it returns the most probable value with lowest potential of being exceeded

shaman · May 24, 2016, 08:38

Anyhow.. fluent has a function of fixed velocity... is this applicable

ghorrocks · May 24, 2016, 21:06

I do not know what the fluent fixed velocity function means. Please explain what it does.

shaman · May 24, 2016, 21:09

It fixed the velocity in a particular domain.... feedback from support.. need to verify this and seek equivalent in cfx

ghorrocks · May 25, 2016, 02:54

If you fix the velocity field then there is no point doing CFD as the velocity is fixed.

You can fix the velocity using a source term in CFX.

shaman · May 25, 2016, 03:03

Thnks sir... wil explore and revert...

mujahidbadshah · May 30, 2016, 07:21

you should use a velocity profile at the inlet boundary using profile BC'S option in CFX. or create a user function for changing inlet velocity against depth. See the post non uniform velocity at inlet.

May 23, 2016, 17:39	Modelling underwater sea current	#1
shaman Member shaman Join Date: Jul 2013 Posts: 37 Rep Power: 13	Hi all, i need to model thermal dilution of cold water discharge into seawater. I have 3 question 1. How to model the seawater current which has a boundary layer like velocity profile... ie max at value A at surface and reduces as it reaches bottom. I have tried to model this as velocity inlet and outlet with same for domain initialization but the steady state analysis converge to a different solution. I need to maintain this current throughout the open sea domain. Can i use source.. if so how to maintain velocity profile. 2. My open sea domain is a hexagon 150m radius .. how to model translational periodics so my body of water represents d sea with underwater current mentioned above. 3. How to introduce scalar variable to track the behavior specific to the cold discharge water

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Current density visualisation (PEM fuel cell add-on module)	pchoopanya	FLUENT	10	August 21, 2023 15:33
PEM fuel cell modelling with and without current collector plates???	pchoopanya	FLUENT	2	October 10, 2013 12:12
how simulating a water-particle current into sea	doronzo	FLUENT	0	September 30, 2011 15:52
Modelling a gaz-liquid counter current flow?	Arou	FLUENT	0	March 9, 2008 07:12
Modelling Jet Flow in a current	Adrian	FLUENT	3	January 12, 2006 11:42

May 23, 2016, 20:49		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,871 Rep Power: 144	1) You will only get your intended velocity when you set the simulation up such that it gives the profile - in other words you have an appropriate turbulence model, turbulence conditions and other variables. So if you don't get the right profile it is because you have not set the model up such that it gives you the profile. 2) Make the matching parallel faces translational periodic pairs. So you will have 3 pairs for your hexagon. 3) Simple way: use an additional variable with convection and diffusion. But this will not account for the variation in material properties due to the difference in composition. If you want to account for this (to model things like the density difference due to salinity content) then use a multicomponent mixture.

May 23, 2016, 20:55		#3
shaman Member shaman Join Date: Jul 2013 Posts: 37 Rep Power: 13	Tqvm.. ghorrocks.. The sea current i m looking to model is less complicated as compared to prevailing sea current with the earth movement..

May 23, 2016, 20:57		#4
shaman Member shaman Join Date: Jul 2013 Posts: 37 Rep Power: 13	I need a body of water moving at a velocity profile of my making vs depth. The profile is from metocean study so it is reasonably accurate... and below 0.8ms-1.

May 23, 2016, 20:59		#5
shaman Member shaman Join Date: Jul 2013 Posts: 37 Rep Power: 13	Isnt it possible to introduce a source term to that body of water... n the discharge water resolved as normal

May 24, 2016, 00:57		#6
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,871 Rep Power: 144	Yes, you can apply a source term. But if you apply a source term everywhere it will probably interfere with the flow around the region of interest to you. It could be better to put the source term on the periodic face only to pull the velocity to the profile. If you know the velocity profile then why not forget about the periodic faces and just use an inlet with a prescribed velocity profile? Then you can specify the the profile easily.

May 24, 2016, 01:46		#7
shaman Member shaman Join Date: Jul 2013 Posts: 37 Rep Power: 13	Without source term wont the velocity at the discharge location diffuse due to large seabody. I could initialize the seabody with similar velocity profile. With no slip floor

May 24, 2016, 01:50		#8
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,871 Rep Power: 144	If you have the correct turbulence model and conditions then the flow should not diffuse too much. This was my first point in the first post - you should try to make the physical model correct first, and then the need for a source term to enforce the desired velocity profile is greatly reduced.

May 24, 2016, 02:30		#9
shaman Member shaman Join Date: Jul 2013 Posts: 37 Rep Power: 13	Only have metocean data as ref for velocity profile.. have to make do with source term .. can the source term be velocity with function of height....and initialize sea body similarly

May 24, 2016, 07:26		#10
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,871 Rep Power: 144	I do not understand what "metocean data as ref for velocity profile" means. Yes, a source term can be a function of height (or just about anything else for that matter).

May 24, 2016, 08:38		#11
shaman Member shaman Join Date: Jul 2013 Posts: 37 Rep Power: 13	Metocean is d oceanographic study.. simulation conducted to simulate design criteria for specific return period ranging from 10... 50... 100 and in between.... it returns the most probable value with lowest potential of being exceeded

May 24, 2016, 21:06		#13
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,871 Rep Power: 144	I do not know what the fluent fixed velocity function means. Please explain what it does.

May 24, 2016, 21:09		#14
shaman Member shaman Join Date: Jul 2013 Posts: 37 Rep Power: 13	It fixed the velocity in a particular domain.... feedback from support.. need to verify this and seek equivalent in cfx

May 25, 2016, 02:54		#15
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,871 Rep Power: 144	If you fix the velocity field then there is no point doing CFD as the velocity is fixed. You can fix the velocity using a source term in CFX.

May 25, 2016, 03:03		#16
shaman Member shaman Join Date: Jul 2013 Posts: 37 Rep Power: 13	Thnks sir... wil explore and revert...

May 30, 2016, 07:21		#17
mujahidbadshah Member Mujahid Badshah Join Date: Oct 2015 Posts: 32 Rep Power: 11	you should use a velocity profile at the inlet boundary using profile BC'S option in CFX. or create a user function for changing inlet velocity against depth. See the post non uniform velocity at inlet.