(PLEASE HELP) Error Y+ adaption

sincity · April 23, 2011, 05:23

i'm simulating butterfly valve with pipe. my simulation has converged but i cannot adapt it (i use y+/y* adaption) because i get an error message warning ST malloc : out of memory resize storage on thread. what should i do guys maybe anyone have same problem with me and give me a solution..please share in here. it's really urgent friend!

i use single computer ( dual core , 4GB of RAM, 1 tera of harddisk)

samsi · April 23, 2011, 06:34

how many elements does your mesh has?I think it's possible that you don't have enough RAM.

sincity · April 23, 2011, 07:43

my initial mesh number is just 70.000 and after i adapt it 3 times then the cell increase significantly..it's about 1.150.000 cell. what should i do?

sincity · April 24, 2011, 07:39

somebody please help!!!do not just read please!

meangreen · May 5, 2011, 14:37

sincity, 1,500,000 cells is a lot. You may be at the upper limit of what your computer can handle. I would suggest trying to re-mesh in gambit. Use the boundary layer options to make the boundary finer near the wall, but not everywhere. That way you may be able to get the y+ you are looking for, but not max out your memory.

the other thing you should try is running the code on a computer with more memory. I have dual core with less RAM and I can run simulations on the order of 500,000 but haven't gone to O(10^6).

first re-mesh in gambit, get the cell size as small as you need just at the wall.

sincity · May 10, 2011, 04:32

my model has curvature geometry and a couple of small gaps its hard to make a boundary layer...if i could i would..

April 23, 2011, 05:23	(PLEASE HELP) Error Y+ adaption	#1
sincity Member wibi ario Join Date: Feb 2011 Posts: 46 Rep Power: 15	i'm simulating butterfly valve with pipe. my simulation has converged but i cannot adapt it (i use y+/y* adaption) because i get an error message warning ST malloc : out of memory resize storage on thread. what should i do guys maybe anyone have same problem with me and give me a solution..please share in here. it's really urgent friend! i use single computer ( dual core , 4GB of RAM, 1 tera of harddisk)

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Mesh Adaption in CFX.	Kushagra Mittal	CFX	1	August 31, 2011 08:17
Adaption Error	Ken Adams	FLUENT	0	July 24, 2007 22:29
boundary adaption with geometry based adaption	Kate	FLUENT	0	October 9, 2005 10:44
On Gradient Grid Adaption	David	FLUENT	2	May 2, 2005 15:50
Need help with Grid Adaption	_stephen_	FLUENT	0	January 13, 2005 07:07

April 23, 2011, 06:34		#2
samsi New Member Join Date: Apr 2010 Posts: 13 Rep Power: 16	how many elements does your mesh has?I think it's possible that you don't have enough RAM.

April 23, 2011, 07:43		#3
sincity Member wibi ario Join Date: Feb 2011 Posts: 46 Rep Power: 15	my initial mesh number is just 70.000 and after i adapt it 3 times then the cell increase significantly..it's about 1.150.000 cell. what should i do?

April 24, 2011, 07:39		#4
sincity Member wibi ario Join Date: Feb 2011 Posts: 46 Rep Power: 15	somebody please help!!!do not just read please!

May 5, 2011, 14:37		#5
meangreen Member Casey Join Date: Jun 2009 Posts: 98 Rep Power: 16	sincity, 1,500,000 cells is a lot. You may be at the upper limit of what your computer can handle. I would suggest trying to re-mesh in gambit. Use the boundary layer options to make the boundary finer near the wall, but not everywhere. That way you may be able to get the y+ you are looking for, but not max out your memory. the other thing you should try is running the code on a computer with more memory. I have dual core with less RAM and I can run simulations on the order of 500,000 but haven't gone to O(10^6). first re-mesh in gambit, get the cell size as small as you need just at the wall.

May 10, 2011, 04:32		#6
sincity Member wibi ario Join Date: Feb 2011 Posts: 46 Rep Power: 15	my model has curvature geometry and a couple of small gaps its hard to make a boundary layer...if i could i would..