Alternating direction implicit (ADI) in square

chberbelali · February 12, 2022, 15:39

Hello everyone,

I'm in the process of programming ADI method (with Thomas algorithm) for a problem of free convection in a wavy square enclosure, doing the transition from physical domain to computational domain, but my code dosen't support a big range of undulation amplitude and mesh refinement.

Please help!

Continuum · February 13, 2022, 04:34

I’d suggest starting with simple heat equation, no flow, fixed source in order to verify your solver. Then move to viscous Burgers equation, and onward. How are you treating advection terms in ADI and pressure solution?

chberbelali · February 13, 2022, 05:51

My governing equations composed by two, one for temperature the other for fluid flow, i didn't use Burgers equation!!

FMDenaro · February 13, 2022, 05:59

Quote:

Originally Posted by chberbelali

Hello everyone,

I'm in the process of programming ADI method (with Thomas algorithm) for a problem of free convection in a wavy square enclosure, doing the transition from physical domain to computational domain, but my code dosen't support a big range of undulation amplitude and mesh refinement.

Please help!

Could you better detail your question? I am not sure about your issue, is a problem of mapping the two domains? Do you have an implicit time step for the diffusion only in vertical direction?

chberbelali · February 13, 2022, 06:34

I'm already tested the code in simple case of square, it can't support a large number of iterations (mesh grids), but when i transfom my new domain to make it square it dosn't perfom better.

The new domain: square cavity with a right wavy wall and heated from the left side.
https://images.app.goo.gl/JDDbEFvEvu4ZjTTz9

my scheme:
https://images.app.goo.gl/AshjXZJvs39YHsQa6

February 12, 2022, 15:39	Alternating direction implicit (ADI) in square	#1
chberbelali New Member Ammar Join Date: Jan 2022 Posts: 11 Rep Power: 4	Hello everyone, I'm in the process of programming ADI method (with Thomas algorithm) for a problem of free convection in a wavy square enclosure, doing the transition from physical domain to computational domain, but my code dosen't support a big range of undulation amplitude and mesh refinement. Please help!

February 13, 2022, 04:34		#2
Continuum New Member Continuum Join Date: Aug 2009 Posts: 19 Rep Power: 16	I’d suggest starting with simple heat equation, no flow, fixed source in order to verify your solver. Then move to viscous Burgers equation, and onward. How are you treating advection terms in ADI and pressure solution? aerosayan likes this.

February 13, 2022, 06:34		#5
chberbelali New Member Ammar Join Date: Jan 2022 Posts: 11 Rep Power: 4	I'm already tested the code in simple case of square, it can't support a large number of iterations (mesh grids), but when i transfom my new domain to make it square it dosn't perfom better. The new domain: square cavity with a right wavy wall and heated from the left side. https://images.app.goo.gl/JDDbEFvEvu4ZjTTz9 my scheme: https://images.app.goo.gl/AshjXZJvs39YHsQa6 Last edited by chberbelali; February 14, 2022 at 07:42.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Advective BC with specified direction (outlet-velocity)	uckmhnds	OpenFOAM Running, Solving & CFD	0	August 19, 2018 20:50
ERROR: Flow direction on the boundaries must not be tangential to the boundary.	turbomax	CFX	2	August 6, 2015 11:05
Changing inflow velocity direction deteriorates lift and drag	ziggo	FLUENT	3	July 24, 2013 08:39
stability of alternate direction implicit scheme	thegodfather	Main CFD Forum	0	September 21, 2012 05:41
ADI of N-S equations	Zhou Yongcheng	Main CFD Forum	5	January 18, 2001 07:58

February 13, 2022, 05:51		#3
chberbelali New Member Ammar Join Date: Jan 2022 Posts: 11 Rep Power: 4	My governing equations composed by two, one for temperature the other for fluid flow, i didn't use Burgers equation!!