Error:: Floating point overflow

arunraj · September 3, 2011, 11:50

Hi,
My Problem is i have to find the exit velocity components of a distortion screen and have to give that as a inlet boundary condition for a complete rotor domain..

I have created the mesh for 90 degree distortion screen through Icem cfd and I have imported that in Ansys cfx....Earlier it has shown i m having isolated volumes ..But i m sure there are no isolated volumes ..So i have set the expert parameter to uncheck the isolated volumes ... After that it has shown that integer and real memory are low..so i have increased the stack memory size..Now finally it is showing "floating point overflow error "...( ONly 10 iterations are going after that solver gets stopped)

type::steady state simulation
Inlet boundary condition :: Cartesian velocity components
u= 100 m/s
v=0
w=0
Outlet boundary condition:: mass flow rate
mass flow rate= 2 kg/s
Remaining faces of the domains :: I have considered it as opening
Opening boundary condition values::
Relative pressure = 1010325 pa
Temperature :: 300k
FOR SCREEN :: I HAVE CONSIDERED IT AS A WALL
For the screen which i have kept inside the center part of the domain, i HAVE CONSIDERED THE SCREEN AS A WALL
Turbulence model:: Shear transport(SST)
Heat transfer model:: Total energy
Solver settings::
Advection scheme:: High resolution
Auto time scale:: conservative:: Time scale=0.75

I am not understanding whether i m doing any mistake in Boundary conditions..So plz help me in sorting this problem

I m attaching a picture of model for reference...
Your help will be highly appreciated

ghorrocks · September 3, 2011, 21:19

I would keep the check for isolated volumes. Your logic for removing it seems bizarre.

Solver overflow means you need to improve the numerical stability of the solution. Smaller time steps, better mesh, double precision are the main suggestions.

I should write an FAQ on this, getting a bit sick of writing this explanation.

arunraj · September 4, 2011, 04:20

Hii Glenn Horrocks,

I am sure there is no isolated volumes thats y I have unckeck it..May be I have poor mesh quality..The problem is I have mesh this geometry in Icem CFD...But can u check and say whether there is any thing wrong with the boundary conditions which i have given?

zkdkeen · September 4, 2011, 07:50

Quote:

Originally Posted by arunraj

Relative pressure = 1010325 pa

Should it be 101325pa?

arunraj · September 4, 2011, 12:22

Hi zkdkeen,
Sry sry its 101325 pa only..Mistakenly typed...But I m thinking now since it is relative pressure I have to give as 0 pa rite...

ghorrocks · September 4, 2011, 21:06

You have a specified velocity inlet and a mass flow outlet. This is poorly posed boundary conditions - read the documentation about defining boundary conditions. You should make one or other a pressure boundary.

sheikh nasir · February 8, 2012, 02:59

Hello.
I am getting floating error:invalid number. I am working on train moving in tunnel. Can any body help me. My email is sheikhnasir39@gmail.com
thanks

ghorrocks · February 8, 2012, 05:58

I think there are four active threads which are asking about floating point errors. This is getting a little tedious.....

Do a search on the forum and find threads talking about this. It has been asked thousands of times.

UPDATE: I have deleted the many duplicate posts you put up on exactly the same topic. Please do not put duplicate posts up, I will just delete them.

September 4, 2011, 04:20	Error:: Floating point overflow	#3
arunraj Senior Member Arun raj.S Join Date: Jul 2011 Posts: 210 Rep Power: 16	Hii Glenn Horrocks, I am sure there is no isolated volumes thats y I have unckeck it..May be I have poor mesh quality..The problem is I have mesh this geometry in Icem CFD...But can u check and say whether there is any thing wrong with the boundary conditions which i have given?

September 4, 2011, 21:06		#6
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,871 Rep Power: 144	You have a specified velocity inlet and a mass flow outlet. This is poorly posed boundary conditions - read the documentation about defining boundary conditions. You should make one or other a pressure boundary. zkdkeen likes this.

February 8, 2012, 02:59	Floating error	#7
sheikh nasir Member Join Date: Jan 2012 Posts: 58 Rep Power: 14	Hello. I am getting floating error:invalid number. I am working on train moving in tunnel. Can any body help me. My email is sheikhnasir39@gmail.com thanks

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
block-structured mesh for t-junction	Robert@cfd	ANSYS Meshing & Geometry	20	November 11, 2011 05:59
Floating point exception: Overflow?	mike	CFX	4	December 15, 2009 18:30
Error #001100279 (floating point overflow)	Reza	CFX	4	September 24, 2009 16:15
[Gmsh] Gmsh and samplesurface	touf	OpenFOAM Meshing & Mesh Conversion	2	December 10, 2007 03:27
CFX4.3 -build analysis form	Chie Min	CFX	5	July 13, 2001 00:19

September 3, 2011, 21:19		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,871 Rep Power: 144	I would keep the check for isolated volumes. Your logic for removing it seems bizarre. Solver overflow means you need to improve the numerical stability of the solution. Smaller time steps, better mesh, double precision are the main suggestions. I should write an FAQ on this, getting a bit sick of writing this explanation.

September 4, 2011, 12:22		#5
arunraj Senior Member Arun raj.S Join Date: Jul 2011 Posts: 210 Rep Power: 16	Hi zkdkeen, Sry sry its 101325 pa only..Mistakenly typed...But I m thinking now since it is relative pressure I have to give as 0 pa rite...

February 8, 2012, 05:58		#8
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,871 Rep Power: 144	I think there are four active threads which are asking about floating point errors. This is getting a little tedious..... Do a search on the forum and find threads talking about this. It has been asked thousands of times. UPDATE: I have deleted the many duplicate posts you put up on exactly the same topic. Please do not put duplicate posts up, I will just delete them.