leastSquares vs. Gauss linear

wildfire230 · January 11, 2019, 11:10

Hi all,

Quick question: is there any general rule about when to pick leastSquares vs. Gauss linear for Grad(U)? From what I've read, it seems like leastSquares is more accurate when you have skewness in the mesh.

Thanks!

akidess · January 11, 2019, 11:42

Quick answer: Always use LSQ unless you run into troubles.

wildfire230 · January 11, 2019, 11:43

Great, thanks so much!

wildfire230 · January 11, 2019, 14:59

Hi, sorry one more question, does that also apply to Grad(p)?

pete20r2 · January 12, 2019, 09:24

I've seen it recommend to use LSQ unless you're on a tet mesh because of osciillitatory behaviour. I've also seen it recommend to never apply limiters on p field and that p should use Gauss linear as it is not a convective term.

akidess · January 14, 2019, 08:51

Why would it matter if the gradient you are evaluating is part of the convective term or not?

wildfire230 · January 14, 2019, 09:20

@akidess: So do you always use leastSquares for both grad(U) and grad(P)? I had one experience recently where a low-Reynolds number solution converged with grad(p) Gauss linear but diverged with grad(p) leastSquares. I guess this is probably due to a poorly designed mesh?

Thanks!

pete20r2 · January 14, 2019, 09:54

Quote:

Originally Posted by akidess

Why would it matter if the gradient you are evaluating is part of the convective term or not?

Not actually sure see:
gradient schemes

January 11, 2019, 11:10	leastSquares vs. Gauss linear	#1
wildfire230 Senior Member Join Date: Jul 2013 Posts: 124 Rep Power: 13	Hi all, Quick question: is there any general rule about when to pick leastSquares vs. Gauss linear for Grad(U)? From what I've read, it seems like leastSquares is more accurate when you have skewness in the mesh. Thanks!

January 11, 2019, 11:42		#2
akidess Senior Member Anton Kidess Join Date: May 2009 Location: Germany Posts: 1,377 Rep Power: 30	Quick answer: Always use LSQ unless you run into troubles. wildfire230, Raphael_Santos and BlnPhoenix like this. __________________ On twitter @akidTwit Spend as much time formulating your questions as you expect people to spend on their answer.

January 12, 2019, 09:24		#5
pete20r2 Senior Member Peter Baskovich Join Date: Jul 2014 Posts: 127 Rep Power: 12	I've seen it recommend to use LSQ unless you're on a tet mesh because of osciillitatory behaviour. I've also seen it recommend to never apply limiters on p field and that p should use Gauss linear as it is not a convective term. Kummi likes this.

January 14, 2019, 08:51		#6
akidess Senior Member Anton Kidess Join Date: May 2009 Location: Germany Posts: 1,377 Rep Power: 30	Why would it matter if the gradient you are evaluating is part of the convective term or not? __________________ On twitter @akidTwit Spend as much time formulating your questions as you expect people to spend on their answer.

January 14, 2019, 09:20		#7
wildfire230 Senior Member Join Date: Jul 2013 Posts: 124 Rep Power: 13	@akidess: So do you always use leastSquares for both grad(U) and grad(P)? I had one experience recently where a low-Reynolds number solution converged with grad(p) Gauss linear but diverged with grad(p) leastSquares. I guess this is probably due to a poorly designed mesh? Thanks! ugurtan666 likes this.

January 11, 2019, 11:43		#3
wildfire230 Senior Member Join Date: Jul 2013 Posts: 124 Rep Power: 13	Great, thanks so much!

January 11, 2019, 14:59		#4
wildfire230 Senior Member Join Date: Jul 2013 Posts: 124 Rep Power: 13	Hi, sorry one more question, does that also apply to Grad(p)?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
p_rgh initial residual no change with different settings	manuc	OpenFOAM Running, Solving & CFD	3	June 26, 2018 16:53
2nd Order Convergence Problem for 3D Airfoil	turkmengokce	OpenFOAM Running, Solving & CFD	1	September 10, 2015 08:20
A turbulent test case for rhoCentralFoam	immortality	OpenFOAM Running, Solving & CFD	13	April 20, 2014 07:32
how to modify fvScheme to converge?	immortality	OpenFOAM Running, Solving & CFD	15	January 16, 2013 14:06
solution diverges when linear upwind interpolation scheme is used	subash	OpenFOAM	0	May 29, 2010 02:23