[jonior]

Divergence detected in AMG solver: x-momentum

this message appears when the calculations begain to start and then it stopped

[LuckyTran]

It is a general divergence error. Usually associated with a floating point error of some sort. Can't really say what caused the error, it could be many things. Check over your settings and make sure everything is set up properly..

The algebraic multigrid solver (AMG) is a subroutine that is run every outer iteration where the restriction and prolongation of the solution onto coarser grids is performed to increase the propagation speed of changes in the solution throughout the domain.

[vasava]

Try hybrid initialization.

[pakk]

Originally Posted by vasava

Try hybrid initialization.

Although this is without doubt well-intended, it probably is not useful advice.

As LuckyTran said, this error message is a very general error message. It can be caused by many different things, and all these things require a different solution. Hybrid initialization could be a solution in some cases, but not in all.

Instead of listing random solutions, I think it is better to give jonior some diagnostic methods, and corresponding solutions.

For example: try to find out where in your domain the solution started to diverge. You can do that by plotting the last solution that you had. Typically, extremely large values appear in some places. (Extremely large means clearly unphyiscal, so not a temperate of 400 degrees but a temperature of 1e20 K, or a pressure of 1e20 Pa.) Do you see anything strange at this location? Is the mesh bad here? If so, it is a sign you should improve your mesh. Is it close to a location where a chemical reaction takes place? If so, something might be wrong with that reaction. And so on.

[vasava]

Originally Posted by pakk

Although this is without doubt well-intended, it probably is not useful advice.

Rightly said. I was in bit of a hurry when I wrote that. Anyways I too would look at my mesh, also plot contours of the quantity in question and try to locate the region where it is abnormal.

[bluenail]

in this case,,,,,

When AMG error, finding error is really hard

Bcz there are many factor affect the AMG solver.

so please check your boundary carefully again

[jonior]

great thanks for every one for all your supports i try to follow your advices and i am checking now every parameter one of them in boundary condition i want to ask about if i have an opening (not pressure in or out) just opening i want from the simulation to predict if the flow will goes out from it or it will be like a vacuum ??? which boundary condition should i use in fluent (note: in CFX boundary condition has a type called opening i didn't found the same in fluent)