CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   Newtons Method failed to converge in 100 iteration (http://www.cfd-online.com/Forums/cfx/22265-newtons-method-failed-converge-100-iteration.html)

 hagupta March 6, 2006 12:13

Newtons Method failed to converge in 100 iteration

I get the error "Newton's Method failed to converge in 100 iterations...." for "Temperature" with location "Vertices".

The error is displayed a few times and soon the solver stops.

How do I go about debugging this?

 Ogbeni March 7, 2006 16:56

Re: Newtons Method failed to converge in 100 itera

Newton's method ? sounds like a Finite Element problem, this is CFD site.

 hagupta March 8, 2006 08:26

Re: Newtons Method failed to converge in 100 itera

Hi Ogbeni,

I know this is a cfd site and CFX forum. I guess every cfx user must have encountered this error at some point when using the CFX Solver.

The exact error goes like: ****** Notice ****** | | Newtons method failed to converge in 200 iterations. This | | occured while computing the following variable: | | | | Variable Name : Temperature | | Location Name : FluidZone | | Mesh location : VERTICES | | Mesh entity : | | Last 3 Changes : 4.59189E+01 4.59189E+01 4.59189E+01 | | Tolerance : 1.0000E-02 | | | | The Newton iteration was either slowly converging or has stalled. | | The solver will continue with the variable set as it was on the | | final iteration. If this situation continues you might try | | increasing the number of iterations allowed for Newtons method. | | This can be changed by setting one of the parameters: | | | | Temperature : "Constitutive Relation Iteration Limit" | | Pressure : "Newton Pressure Iteration Limit" | | | | for your mixture using the definition file editor. | +----------------------------------------------------

 Joe March 10, 2006 03:28

Re: Newtons Method failed to converge in 100 itera

The Newton iteration is used for calculation of thermodynamic properties of mixtures. How is your physical setup? What kind of fluid do you have? Without answers to that questions it is not possible to help you.

 hagupta March 10, 2006 10:36

Re: Newtons Method failed to converge in 100 itera

Thanks Joe. I had no idea that it is related to thermodynamic properties of mixtures.

Fluid: Mixture of 3 user-defined gases.(Variable Composition Mixture). [Cp(T),Thermal cond(T), Viscosity(T) have been defined as functions of temperature apart from the mol. wt. and other essential parameters]

The setup basically consists of a cylinder with 3 inlets, one for each of the gases and an outlet.

As I reduce the timestep(at the beginning of a run), the no. of iterations after which this error comes(and stops the solver) increases. However, for a very small timestep of the order of -5(approx. 1/10th of the autotimestep value) , this error seems to have vanished.

Could this be because of wrongly defining the properties of the gases? or anything to do with the timestep? or anything else?

 Joe March 10, 2006 12:20

Re: Newtons Method failed to converge in 100 itera

Yes, it could be that there is a mistake in your property defintion. Especially the cp values, as the problem is related to temperature. I would check them. If you have a result file or backfile make a chart cp over T.

 Stone March 13, 2006 06:01

Re: Newtons Method failed to converge in 100 itera

I also got this error message although I'm using CFX materials. A binary mixture of O2 and JetA. Cp is given in Nasa Format for both materials.

Does anybody have an idea what causes this error message?

 Joe March 18, 2006 07:30

Re: Newtons Method failed to converge in 100 itera

JetA is that part of the cfx material properties data bases or is it your own data? If its your own data, check first the simple things like typos, dimensions etc.

 Reza_computational October 31, 2010 18:08

newton method

I am simulating multi-phase flow of water injected into hot air stream and the water is supposed to evaporate. the materials are:
1.a variable mixture of air and water vapour
2.a variabl mixture of liquid and gas phase water
air is continuous and water mixture is dispersed fluid. Interfacial equilibrium nodel is Raoults law. CFX-Pre settings are according the instructions in help. there is no errors in CFX-Pre. but while solving after a few iteration the temperature starts diverging reaching to 8e30 K. CFX solver prompts in every iteration that the newton method has been diverged after 100 iteration and user needs to change pressure or temperature underrelaxation in definition file editor (which i also dont know where it is)...
appreciate ur help through this mess!
Reza

 leomasc December 2, 2010 07:41

+--------------------------------------------------------------------+
| ****** Notice ****** |
| Newtons method failed to converge in 150 iterations. This |
| occurred while computing the following variable: |
| |
| Variable Name : Total Pressure |
| Location Name : inlet |
| Mesh location : BELG6 |
| Mesh entity : VERTEX |
| Last 3 Changes : 2.42581E+01 2.42581E+01 2.42581E+01 |
| Tolerance : 1.0000E+00 |
| |
| The Newton iteration was either slowly converging or has stalled. |
| The solver will continue with the variable set as it was on the |
| final iteration. If this situation continues you might try |
| increasing the number of iterations allowed for Newtons method. |
| This can be changed by setting one of the parameters: |
| |
| Temperature : "Constitutive Relation Iteration Limit" |
| Pressure : "Newton Pressure Iteration Limit" |
| |
| for your mixture using the definition file editor. |
+--------------------------------------------------------------------+
| H-Energy | 1.01 | 3.6E-04 | 2.4E-02 | 5.5 3.5E-02 OK|
+----------------------+------+---------+---------+------------------+
| K-TurbKE | 0.99 | 1.9E-04 | 6.4E-03 | 5.5 1.5E-02 OK|
| E-Diss.K | 1.00 | 4.3E-04 | 1.5E-02 | 7.2 9.5E-04 OK|
+----------------------+------+---------+---------+------------------+
+--------------------------------------------------------------------+
I got this message when simulating the flow in exhaust of a steam turbine .The matiral used was wet steam.
It became very hard to converge when the notice appear.
Does anybdoy have a idea to handle it?

 All times are GMT -4. The time now is 14:23.