[GV77]

Hi.

I'm modelling a burner mixer to burn NH3 and H2 and I have questions about how could I mesh my domain.
The domain is simple: products from the first stage enter on the left and then there are two jets, the one having a smaller diameter corresponding to the H2 jet and the other one corresponding to the air jet. Both these jets enter in the domain and mix with the cross flow, and then they possibly react with each other.

The question that I have is that if I could divide the domain, detaching the jets from the mixer in order to create 3 diferent meshes: 1 for each jet and other for the burner mixer fluid volume. The problem is that if I do so, even if I include these 3 meshes in the same region, Star CCM would create interfaces in the jets junctions with the mixer. And those areas are in somehow critical because it's where there are steep gradients and I don't know if Star CCM interfaces can solve them correctly, without convergence problems or another issues.
By the way, I confirmed and the flow is incompressible everywhere in the domain.

May someone give me an opinion about this?

[cwl]

If you're in doubt about effect of mesh at interfaces on the convergence - Star can generate conformal mesh at interfaces.

[GV77]

Ok, but when you say conformal mesh, you say that I need to create contacts that are conformal or I should get a conformal match between interfaces?
I know that sounds a little weird because supposedly I can only create conformal meshes/interfaces with conformal contacts.

But for example, in this case I applied a contact between the two sides of the domain. As the domain is a burner slice and I splitted it into 3 subdomains: burner upperstream (premix), burner middle stage (mix) and burner downstream (lean stage) in reality there are 6 side boundaries, 3 from each side (images attached). I defined them as periodic contacts in order to create periodic boundaries. Premix and Lean Stage are meshed with directed meshing (reduce computational costs), so even though I applied periodic contacts on the left and right side of these blocks, I couldn't create a conformal contact. However when I activate the interfaces, I can create a conformal match between them (because the mesh surfaces match).

That's why I'm asking if it's mandatory conformal contact creation or I only need to get the conformal match confirmation in the output window.

[gustavson]

I sadly do not have an answer for you but rather a question. How did you manage to visualize the velocities of all your components. Do you have multiple regions and an aquivilant number of physics? How did you connect these.
I am trying to visualize the mixing of two seperated liquids that imping on each other and their resulting velocity. I tried 3 regions with a liquid physics but star ccm connects them as baffle interfaces.

[GV77]

In terms of visualization, I created a scene (vector scene) where I created two (vector) displays: a display to visualize the velocity and another to visualize the direction vector.
For the velocity display I changed display mode from gliph to line integral convolution and the opacity to 0.896 (or another value diferent of 1 because if you don't do that, you can't see the direction vector very well).
For the direction vectors, I just created a vector display and defined a vector field called "Flow Direction General" (star doesn't have this field by default) to see the adimensional velocity vector field (adimensionalized by the velocity magnitude).

Answer to the second question: I have created three mesh blocks but I've assigned only one region to them and only one physics. I haven't separated the domain in the star ccm+. I splitted it using Solidworks. And when I imported the geometry from solidworks, starccm+ imported 3 parts: mix, premix and lean stage. Then I assigned those parts to 1 region, but I meshed the parts diferently (as I've explained in the previous post). I think that the only thing that you need to do to create a interface like my interfaces, you need to go to the interfaces, click in the interface you want to change (baffle interface) and then change (interface) type from baffle interface to internal interface and topology(right below type) to in-place if it's not already assigned to it.
I don't know if you have created three parts or if you have only one part. If you only have one part, you don't need to create interfaces or contacts. But if you have two or three parts connected to each other, even if they are assigned to only one region, I think you must create interfaces inside of the region.

P.S.: I'm not a Star/CFD expert or specialist, so my advices may be right or not. They are questionable.

[gustavson]

Thank you for your help. Using the Multi-Component Liquid Model worked for me and my simulation works as intended. I previously tried several multi-phase models but couldnt quiet get the inflow settings right. I also tried with different numbers of parts.
I am simulating an injector and I created 3 parts with my fuel, ox and chamber region. I have set the interfaces as internal interface and it works fine.