Residual spike while switching form serial to parallel.

nvarma · May 26, 2016, 11:41

Can anyone explain why there is a spike in residuals as I switch (write cas and dat, open in parallel) from serial to parallel run on fluent?

nvarma · May 27, 2016, 11:12

Update:

So not only is there a residual spike, it diverges immediately. The simulation can go on for thousands of iterations in serial but not even hundred on parallel.

Anybody?

`e` · May 27, 2016, 19:27

Could you show us your mesh with the partitions outlined? Be careful of having partition boundaries where important flow characteristics occur (non-linear changes in velocity); especially if you're using a second order discretisation scheme (partitions only overlap one cell on the partition boundary). Have you tried running in parallel from initial conditions or only from a serial-produced result?

May 26, 2016, 11:41	Residual spike while switching form serial to parallel.	#1
nvarma Senior Member nm Join Date: Mar 2013 Posts: 100 Rep Power: 13	Can anyone explain why there is a spike in residuals as I switch (write cas and dat, open in parallel) from serial to parallel run on fluent?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Maximum number of iterations exceeded chtmultiregionsimpleFoam	Moncef	OpenFOAM Running, Solving & CFD	28	July 13, 2020 15:26
time step continuity problem in VAWT simulation	lpz_michele	OpenFOAM Running, Solving & CFD	5	February 22, 2018 20:50
simpleFoam error - "Floating point exception"	mbcx4jc2	OpenFOAM Running, Solving & CFD	12	August 4, 2015 03:20
Wrong fluctuation of pressure in transient simulation	caitao	OpenFOAM Running, Solving & CFD	2	March 5, 2015 22:33
Unstabil Simulation with chtMultiRegionFoam	mbay101	OpenFOAM Running, Solving & CFD	13	December 28, 2013 14:12

May 27, 2016, 11:12		#2
nvarma Senior Member nm Join Date: Mar 2013 Posts: 100 Rep Power: 13	Update: So not only is there a residual spike, it diverges immediately. The simulation can go on for thousands of iterations in serial but not even hundred on parallel. Anybody?

May 27, 2016, 19:27		#3
`e` Senior Member Join Date: Mar 2015 Posts: 892 Rep Power: 18	Could you show us your mesh with the partitions outlined? Be careful of having partition boundaries where important flow characteristics occur (non-linear changes in velocity); especially if you're using a second order discretisation scheme (partitions only overlap one cell on the partition boundary). Have you tried running in parallel from initial conditions or only from a serial-produced result?