Learning Forum

[Dubey92]

I am modeling Laser melting of a powder bed. The powder bed consists of powder particles and I have patched the particles with volume fraction of one of the phases. When I do the visualisation of the results in the CFD-Post, the particles are still hollow although their shape changes due to melting. I want to see them as complete partilclles. Here is my domain before it is visualised in CFD-Post and particles are visible hollow.

 

 

Here is an image of a published result and this is how I want to visualise it.

 

Any help is much appreciated. Please revert if someone knows anything about solving this issue.

[SRP]

 

 

[Rob]

How did you patch in hollow particles?

[Dubey92]

Hi Rob. Thanks for your reply. I created spherical cell registers and then patched them with metal volume fraction equal to 0.5 as the domain consists of two phases i.e., metal and gas. I visualised the above by using iso-surfaces of volume fraction equal to 0.5. I might be doing something wrong in visualisation. I am not sure about it.

[Rob]

That's not a hollow sphere, it's a solid sphere with a lower volume fraction. A tennis ball is hollow, you've got a sponge ball. If you're using VOF that's a really bad idea, due to the way the model works. 

[Dubey92]

Thanks for your input. How to correct it? I knew that parching is the option to create the particles from radius and sphere and so I did that.

[Rob]

If you want hollow spheres you'll need to create a second register. Patch the first register as all liquid, then patch the second (smaller diameter spheres) as the hollow bit. I assume the void is air?  Note, as you'll need 5-10 cells across the sphere shell and an aspect ratio of 2-ish your mesh is going to be rather large. 

[Dubey92]

I don't want hollow spheres. I want solid particles only as shown in the second image in my post. I created spherical registers and patched them with metal but I am getting something like the first image, not solid(sponge as you mentioned).

[Dubey92]

The air is between the particles not inside the particles. The particles are purely solid.

[Rob]

Ah, that's not hollow. Where the free surface touches the side of the domain there is no area volume fraction of 0.5 (it's 1.0 liquid).  As you are displaying an isosurface of 0.5 you're seeing a gap/hole. 

[Dubey92]

Yes, I have plotted an iso-surface for volume fraction of metal equal to 0.5 and I am seeing hole inside the particles. Please help how to solve this. My whole simulation result is coming out to be wrong due to this. The second image which I have posted, I am unable to understand how they have done this.

[Rob]

There are no holes. You are seeing a gap in the surface caused by symmetry/wall boundary having a volume fraction that isn't 0.5    The second image may be using volume rendering from CFD Post, or may be using iso-clips. It is not an error, it's because you are displaying a surface at a specific volume fraction. 

[Dubey92]

Thanks Rob. So I assume that when I create spherical cell registers and patch them with the volume fraction of a phase, they are full solid spheres. Right? And the hollowness which I am seeing inside my particles is just a visualisation issue?

[Rob]

A register is a set of cells that meet a criterion. If you used sphere then by default it's solid. 

An iso-surface is just that, a surface. It's "thin" and represents a single value. So, an iso-surface of mesh would be a plane. For the VOF surface above it's the shell: the plastic part of a ping-pong ball. 

[Dubey92]

Thanks Rob. I have used spherical cell registers. So now I am relaxed that I have doing something wrong with visualisation and not with patching. Can you suggest something on how to visualise full solid particles as shown in the second image I have posted?

[Rob]

CFD Post has a solid rendering option, otherwise have a look at iso-clip and plot that and the isosurface.  

[Dubey92]

Okay. Thanks a lot.