Learning Forum

[waqas ahmad]

float point exception.........error in AMG solver..temperature??????? 

please help how to fix it..............

value of current density shoot to 1000+ amps/cm2 after 20 iterations/..............

[Karthik R]

Hello Waqas,

Thank you for your question. I have moved this post to the 'Fluid Dynamics' thread where this will receive more traction. 

In order to answer your question, please share some more details about your model. What are you modeling? How are you model? What are your boundary condition? What is your mesh quality? Please share as many details as possible as there could be multiple reasons you are seeing this issue.

Best Regards,

Karthik

[waqas ahmad]

Modeling :single channel Proton exchange membrane fuel cell 

software Ansys workbench 19.0

mesh type : tetrahedral mesh

gas channel shape : rectangular with triangular head...

skewness 1.0

nodes : 60000

elements 80000

UFRs

Pressure  0.55

momentum 0.3

all other =1

operating pressure 1 atm

temperature : 323 K

isothermal model........ 

BCGstab method 

 

 

 

[Rob]

 I'd recommend reviewing the mesh quality: a cell equiangle skewness of 1.0 essentially means it's flat and has near enough zero volume.  You're aiming for a maximum of 0.8 to 0.9 (ideally under 0.. 

[waqas ahmad]

how can i improve the cell skewness? by reducing number of seeds along the edges? or by changing smoothness factor?????

[Rob]

Yes, no, maybe. You need to look at the mesh and see where the skew cells are then work out why they're skewed and fix it. Common causes are over constrained mesh (you've added too many settings and the algorithm has no way to accommodate all of these)  or too big a jump in size in a narrow gap. 

If you post some images the wider community may be able to offer advice. 

[Karthik R]

Hello,

You might also want to think about decomposing your body appropriately to create a structured grid. Structure flow aligned grid will always improve skewness and aids model convergence. Having said that, it might not always be possible to make a structured grid. In such case, you might want to follow rwoolhou's mesh generation advice.

Best Regards,

Karthik

[waqas ahmad]

I have tried to reduce skewness upto 0.95 .. but issue is still existing .... after 20 iteration current value for anode/cathode shoots up and reaches 100000+ ...then float point error appears....

"""""" amg solver error:temperature""""'

please help

[Karthik R]

Hello Waqas,

Please describe the geometry with screenshots. Please explain your final goal and how you have set-up your model. Please delineate your boundary conditions. Please show a screenshot what your mesh looks like.

Depending on the model, there are multiple reasons why you could be seeing divergence. You have to help us understand your model better so we can help you resolve this issue.

Thank you.

Best Regards,

Karthik

[waqas ahmad]

[waqas ahmad]

images are attached below: 

[waqas ahmad]

[waqas ahmad]

[waqas ahmad]

[waqas ahmad]

[waqas ahmad]

[waqas ahmad]

[waqas ahmad]

[waqas ahmad]

[waqas ahmad]