Why are oceanic models written in primitive variables?

Lupocci · September 16, 2010, 12:17

Hi. I always wandered why all the oceanic models solving the 3D hydrostatic shallow water equations(ROMS,DELF3D,FVCOM)are written in primitive variables. I think the reason is that they do not usually have to deal with shocks, as instead is true for river models in which you can have discontinuos phenomena like dam break or hydraulic jump. Is this the main reason? and if this is the reason, what is the advantage of using primitive variables?
And finally,if this is the reason, these models are not able to compute, for example, a tzunami wave approaching the coast, since the discontinuos wave that forms has the wrong speed (Toro,2001) , hasn't it?
thanks
Alberto

jed · September 20, 2010, 07:39

Shocks are generally thought to be unimportant over long time periods. The methods used are only valid for smooth solutions anyway (this is enforced through artificial dissipation), and are thus not appropriate for purposes like modeling tsunami run-up. I'm not aware of any true advantage of velocity-thickness form, I suspect that choice is mostly historical.

September 16, 2010, 12:17	Why are oceanic models written in primitive variables?	#1
Lupocci New Member Lupo Ci Join Date: Aug 2010 Posts: 16 Rep Power: 15	Hi. I always wandered why all the oceanic models solving the 3D hydrostatic shallow water equations(ROMS,DELF3D,FVCOM)are written in primitive variables. I think the reason is that they do not usually have to deal with shocks, as instead is true for river models in which you can have discontinuos phenomena like dam break or hydraulic jump. Is this the main reason? and if this is the reason, what is the advantage of using primitive variables? And finally,if this is the reason, these models are not able to compute, for example, a tzunami wave approaching the coast, since the discontinuos wave that forms has the wrong speed (Toro,2001) , hasn't it? thanks Alberto

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Incompressible Turbulence models	achinta	OpenFOAM	4	May 27, 2010 10:35
Two-fluid models vs mixture models for bubbly flows	Hansong Tang	Main CFD Forum	6	December 8, 2009 03:21
PHI file structure	Eugene	Phoenics	9	November 2, 2001 22:00
writing a 3d NS solver in the primitive variables!	lostsoul	Main CFD Forum	4	October 20, 2000 02:52

September 20, 2010, 07:39		#2
jed Member Jed Brown Join Date: Mar 2009 Posts: 56 Rep Power: 19	Shocks are generally thought to be unimportant over long time periods. The methods used are only valid for smooth solutions anyway (this is enforced through artificial dissipation), and are thus not appropriate for purposes like modeling tsunami run-up. I'm not aware of any true advantage of velocity-thickness form, I suspect that choice is mostly historical.