[IvanSh]

Hi, everyone!

Could anybody explain where does the error come from, please?

I am trying to simulate a gas metal arc welding process.

I created geometry which consists of two domains separated by a common interface. One domain is defined as multiphase(steel and air) and the other one as a single phase (steel).

The multiphase domain represents the formation of the molten droplets falling on the first side of the interface.

The single-phase domain represents the flow of the solid metal, basically, it represents the plate.

So, the multiphase domain is above the single phase.

I switched off the option "constant domain physics" in order to have multi and single-phase domains respectively. I was thinking maybe that could be the problem, but I am not sure since I don't know the other option to make separate domains.

Hope somebody could clarify this problem. Any tips, suggestions are very welcome.

Thank you in advance.

[Opaque]

Do you have any expression that refers to Temperature?

[IvanSh]

Well, yes. I have defined viscosity-temperature dependence in order to distinguish solid and liquid metal.

Secondly, I am using liquid metal which I defined in materials with reference temperature - Tmelting = 1733K., and thermal conductivity, which I defined in functions.

and, for the molten droplets, I am assigning the expression, which basically says that they have the temperature 2500K and the rest of the multiphase domain is 300K as well as solid metal in the single phase.

[Opaque]

It seems the software is confused about which temperature to use in some expression.

If you can tell which phase the expression must be evaluated for, I would try something along

MyPhase.Temperature

in the expression. Then, the software does not have to guess which phase is implied for the expression. For single phase flows, there is no confusion, but for a domain with a multiphase model it can be a problem.

Keep us posted, and good luck

[IvanSh]

Thank you for your advice!
I found my stupid mistake.
All my expressions were defined with the Prefix "Stahl.T", with a capital "S".
However, in the domain, in the basic settings, the material was defined as "stahl", small "s".
It was not showing any conflict with that, so I was not able to notice this mistake.
But when I finally changed the name of the material to "Stahl," the solver started to work.