[M_Tidswell]

Dear All

I am trying to use ICEM to create a mesh for this swirling combustion can but I am clearly slightly out of my depth. Would someone be kind enough to point me in the direction of how to deal with all the small holes and the swirl vanes, have been banging my head against this for a while and each time it just looks horrible. I think the blocking strategy I'm trying to use is wrong.

Have attached a couple of jpgs of the solid model of the can - the first one is looking into the can from the top and the second is from the bottom.

Many thanks!

Mark

[mezah]

If you are going to use blocking, you can split the block edges around the holes, and make each hole surronded by rectangle or square shape, then delete the bock inside each square to make it hollow or leave it if it is solid, then assigne the curves to the hole.

[M_Tidswell]

Hi There

You say "if your going to use blocking" is there a better way of doing it? Thats the only way I know in ICEM but im more than happy to try something else. I have been trying to do as you suggest but so far it has not been pretty.

Thank you!

[PSYMN]

Yes, you could try using a tetra/prism approach and be done pretty quickly...

If you decide you need the hexa approach, you just need to adjust your thinking a bit. Those turning vanes are twisted geometry, but topologically, they are straight. You would just block them straight in terms of topology and then fit that topology to the curved geometry.

As for the holes, Mezah is right. Here is a bit more detail.

Topologically, this model is several ogrid layers with a chopped up cartesian layer in the middle. So start with a block, and create an Ogrid with faces on the ends. Adjust the blocking so the square is within the central circle, then split for each of the rings. For the holes in the middle square, you would split out a box for each and associate it with the circle. Put the "solid" parts in SOLID. Then, assuming you want flow thru the holes (rather than modeling the solid), you would put an Ogrid thru the whole volume so it passes thru the holes.

It is actually a pretty straight forward model, though those little holes will be tedious. If you can reduce it to a periodic model, you can save some time...

[Far]

Originally Posted by M_Tidswell

Dear All

I am trying to use ICEM to create a mesh for this swirling combustion can but I am clearly slightly out of my depth. Would someone be kind enough to point me in the direction of how to deal with all the small holes and the swirl vanes, have been banging my head against this for a while and each time it just looks horrible. I think the blocking strategy I'm trying to use is wrong.

Have attached a couple of jpgs of the solid model of the can - the first one is looking into the can from the top and the second is from the bottom.

Many thanks!

Mark

Simon explained the hexa mesh in great detail.

but hexa mesh would be the time consuming job. You can get quick results (and mesh too ) with tetra + prism mesh...

[M_Tidswell]

Hi Simon

Thank you for your help, can I just clarify when you say put an "o-grid through the whole volume" do you mean get to the stage you have a box for each hole then associate those box edges to the curves and then put an o-grid through each hole? Yes you are correct the pics there are the solid bits so I would want fluid through any free area.

Many thanks

Mark

[PSYMN]

No, you box out and associate with the holes... but when it comes time to put in the Ogrid, you can select all the bodies (not just the ones in the holes), and put faces on the ends of the model and create your Ogrid...


Alternatively, if you just wanted the Ogrid closer to the holes (to capture a jet of fluid passing thru), you could select the blocks in the holes, plus the blocks in the fluid on either side, no faces, and apply. (you can select these 3 blocks for all the holes at once and apply the Ogrid at once so it is easier to adjust the edge distributions later)

If you only set Ogrid on the blocks in the holes, you will have very poor quality because the outside of the Ogrid would be constrained to the curves resulting in 180 degree angles in those 8 corner elements...

[M_Tidswell]

Hi Simon

Thank you very much, I was not aware you could block everything out and then put the o-grid through all the bodies. I have only done external flows with ICEM in the past so that's a new trick to me.
Can I impose one last question on you - when you block out the curved vanes I assume you delete the blocks that contain the solid vanes. Same logic as the holes, I can then put the o-grid through the whole lot at the end. I don't need to do anything else.

Thank you very much for your time

Mark

[PSYMN]

Right, you can delete the solid blocks or put them in a Solid material and just turn it off. Then when you place the final boundary layer ogrid, it will also pour thru those guide vane holes... Just like a fluid filling the whole volume.

[M_Tidswell]

Dear Simon

I tried to follow your instructions and thought I was getting somewhere but then I discovered that if I didn't put an o-grid through the core to create the vanes I couldn't form all the vanes correctly. Please see pic 1

So I tried putting an o-grid through it first and actually had some success with the vanes after a bit of playing, the only issue I'm having is that the blocks don't seem to want to snap into place but I'm sure that's me being silly somewhere. So I moved on to look at the middle holes and I seem to have made so much of a mess sorting the vanes I'm really not sure where to start with clearing it up so I can think about creating a block for each hole.
Have included a top and a bottom pic of the blocks I'm working on, you can see the non-snapping problem in the top one as well.

I wondered if you could point me in the right direction as to how to start, I tried applying the pre mesh at this stage just to see and that was expectedly horrible too.

Thank you very much

Mark