[pashazanousi]

I simulate a wave maker flume...
But when I define the Gravity in y-direction as a expression an error occur...
It's a simple expression G=0.1*g
I've search web for this error "Fatal overflow in linear solver"...
But I didn't get result...
I'm really confused...
Time step or mesh size is not the problem...
I change Allocate memory from 1 to 1.4...
I need help...
Wave-Maker-Theory.txt

[karachun]

I see that you use VOF. Free surface aligned to partition boundary can cuse divergence.
Try to set solver to serial mode or manualy set partitioning to ensure that free surface is not coincident with partition boundary.
Fatal overflow in linear solver occur when execute solution in parallel

[ghorrocks]

There is an FAQ on this error: https://www.cfd-online.com/Wiki/Ansy...do_about_it.3F

In your case you will need to do some basic debugging.

Are you sure your mesh motion is correct? See FAQ for tips: https://www.cfd-online.com/Wiki/Ansy..._went_wrong.3F

Where did you get your time step size from? I bet you guessed it, and I bet your guess is wrong. Try a smaller time step.

[pashazanousi]

Originally Posted by karachun

I see that you use VOF. Free surface aligned to partition boundary can cuse divergence.
Try to set solver to serial mode or manualy set partitioning to ensure that free surface is not coincident with partition boundary.
Fatal overflow in linear solver occur when execute solution in parallel

Thank you...It solve...

[pashazanousi]

Originally Posted by ghorrocks

There is an FAQ on this error: https://www.cfd-online.com/Wiki/Ansy...do_about_it.3F

In your case you will need to do some basic debugging.

Are you sure your mesh motion is correct? See FAQ for tips: https://www.cfd-online.com/Wiki/Ansy..._went_wrong.3F

Where did you get your time step size from? I bet you guessed it, and I bet your guess is wrong. Try a smaller time step.

No....I checked the mesh...it's ok...Time step is T/200 according to paper....Thank you for your attention

[karachun]

Originally Posted by pashazanousi

Time step is T/200 according to paper....

If you want to be sure - check case with T/400, this will be your timestep convergence study. If nothing changed then you can be sure. Many things like mesh density, domain size, timestep, residuals are varying from problem to problem.
Never thrust papers or folks that say that some setting value is 100% enough. Often you can get advice about good setting values you can start with. Only convergence check (mesh convergence, timestep convergence, residual convergence) can answer this question.

[pashazanousi]

i tried to do a Fluid-Structure Transient Simulation....
But I received the below image error
Untitled.jpg
I do tutorial oscillation plate...But I don't know what this error mean?

[ghorrocks]

Please start a new thread for a new question in future.

The error message is very clear. In the ANSYS Mechanical side of things an element has become highly distorted and the solver failed. You are going to have to do something to stop the excessive distortion. Look at the distorted elements in the ANSYS Mechanical post processor to find the problem element.