Spatial discretization

star · August 20, 2013, 04:35

Hi. I read about the the spatial discretization of momentum and it says that 2nd order upwind gives much accurate result and it is better to get converge solution with first order and then switch to 2nd order and run it till convergence.
So does it mean that we must always use 2nd order upwind method for our final result?
And what about 2nd order Pressure method. Should we also use 2nd order for final result?
Thanks

sbaffini · August 26, 2013, 04:58

As your question is very basic i won't go with details. Just know that 1st order is just not accurate enough and overly dissipative. Second order (or higher if available and really higher, which is not the case for fluent) is thus mandatory. This usually apply also for the time discretization, if unsteady. For the pressure it is the same, but the mentioned 2nd order scheme is not the only second order scheme in Fluent, it is more a "label". Other schemes for pressure, depending on the application, might be suitable as well.

star · August 26, 2013, 07:21

Thanks sbaffini for your reply. You are right. there are different schemes available other than 2nd order but someone suggested me 2nd order for my case. I am simulating 2d airfoil at low Reynold number i.e 1000. The flow is laminar and transient. What do you suggest about pressure, momentum and time discretization about my case to get accurate result?
Thanks again

August 20, 2013, 04:35	Spatial discretization	#1
star Member Harry Join Date: May 2013 Posts: 68 Rep Power: 12	Hi. I read about the the spatial discretization of momentum and it says that 2nd order upwind gives much accurate result and it is better to get converge solution with first order and then switch to 2nd order and run it till convergence. So does it mean that we must always use 2nd order upwind method for our final result? And what about 2nd order Pressure method. Should we also use 2nd order for final result? Thanks

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Spatial vs Angular Discretization in DO radiation model	renaldi	FLUENT	1	February 16, 2024 08:47
ANSYS FLUENT spatial discretization	Akim679	FLUENT	0	January 28, 2012 11:37
Comparison of different spatial and temporal discretization schemes	harly	OpenFOAM Running, Solving & CFD	2	September 24, 2011 12:15
temporal and spatial discretization.	pity	Main CFD Forum	0	June 10, 2007 22:53
Spatial and temporal discretization	Giuseppe	FLUENT	2	January 23, 2006 05:18

August 26, 2013, 04:58		#2
sbaffini Senior Member Paolo Lampitella Join Date: Mar 2009 Location: Italy Posts: 2,152 Blog Entries: 29 Rep Power: 39	As your question is very basic i won't go with details. Just know that 1st order is just not accurate enough and overly dissipative. Second order (or higher if available and really higher, which is not the case for fluent) is thus mandatory. This usually apply also for the time discretization, if unsteady. For the pressure it is the same, but the mentioned 2nd order scheme is not the only second order scheme in Fluent, it is more a "label". Other schemes for pressure, depending on the application, might be suitable as well.

August 26, 2013, 07:21		#3
star Member Harry Join Date: May 2013 Posts: 68 Rep Power: 12	Thanks sbaffini for your reply. You are right. there are different schemes available other than 2nd order but someone suggested me 2nd order for my case. I am simulating 2d airfoil at low Reynold number i.e 1000. The flow is laminar and transient. What do you suggest about pressure, momentum and time discretization about my case to get accurate result? Thanks again