CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   Main CFD Forum (http://www.cfd-online.com/Forums/main/)
-   -   Cutting a hole in the engine cylinder head in KIVA (http://www.cfd-online.com/Forums/main/13024-cutting-hole-engine-cylinder-head-kiva.html)

CWL February 22, 2007 10:57

Cutting a hole in the engine cylinder head in KIVA
 
I am trying to cut a hole on the cylinder head (with KIVA 3V) so that air can be inducted through the hole (as previously told that an injector cannot be used to induct air).

While I can sucessfully designate a cell on the cylinder head as velocity inflow boundary, but, there seems no incoming air flow. I appreciate if anyone can tell me how to cut a hole on the cylinder head.

By the way, what are the difference(s) between velocity and pressure in(out)-flow boundary?

Any advices on these issue will sincerely be appreciated.

Thanks again.

George February 22, 2007 12:04

Re: Cutting a hole in the engine cylinder head in
 
You need more than one cell to simulate the hole.

CWL February 22, 2007 13:35

Re: Cutting a hole in the engine cylinder head in
 
Thanks for replying. Why do I need more than one cell for a hole? By the way, is desinating a bunch of cells with open boundary the way to simulate a hole?

Thanks a lot.

George February 22, 2007 15:05

Re: Cutting a hole in the engine cylinder head in
 
I don't know how KIVA employes the open boundary or if it has any other methods of simulating it. I think outlet pressure boundary is more appropriate. You need more than one cell, because if you try to calculate the gradients, for example the velocity gradients at that cell, it will be zero, instead of having a value.Also two cells will not solve the problem, since you need to "smooth" a little bit of these gradients. If you have du/dx in three cells in a row the values: 0, 2000, 0, then the solver propably will diverge.

CWL February 23, 2007 15:52

Re: Cutting a hole in the engine cylinder head in
 
Thank you very much for reply. One other question, if I have a 3-D cylinder mesh and I want to cut a hole on the top surface (circle face in r-theta plane). Do I have to have a fine 2-D mesh around the hole only on the top r-theta plane? Or, do I have to have a fine 2-D mesh around that area for the full cylinder height? Also, do I have to refine the mesh density in the axial direction?

Your previous reponses was extremely helpful in providing guidelines and insights towards a possible resolution to my current problems in my simulation. I sincerely appreciate and thank you very much for your time and assistances.

By the way, do you also use KIVA for combustion calculations?

George February 24, 2007 09:10

Re: Cutting a hole in the engine cylinder head in
 
--<Also, do I have to refine the mesh density in the axial direction?>

Answering this question depends on your mesh-density at z-direction. If it is quite fine, then you don't have to remesh at the axial-direction. Meaning fine, I mean the dz cell height to be lower than around 2-3 mm. In that case you can leave it as it is.

--<Do I have to have a fine 2-D mesh around the hole only on the top r-theta plane? Or, do I have to have a fine 2-D mesh around that area for the full cylinder height?>

Ofcourse you need a quite fine mesh at the top r-theta plane, but not for the full cylinder height. Only for some grid lines across z-direction (around 5-6). In the case that from these holes fluids enters or leaves with a high velocity, then you should add a couple more.

All the numbers of cells-grid lines that I have mentioned come from experience. There isn't any correct number of cells for that case, but only an approximation. So, don't follow strictly the numbers, but try different configurations and compare the rate of convergence and the results at these difficult to handle areas.

I have dealt lately with an hydrogen-injector placed at the intake manifold of an internal combustion engine, so I had a similar problem. In a way the injector was also a hole and I used the numbers that I have mentioned (intake manifold grid: 18x18x80, cylinder grid: 54x54x40).

--<By the way, do you also use KIVA for combustion calculations?>

I don't use KIVA (I also I don't have it). I am writing my own CFD code (solver+grid generator) for in-cylinder simulations. The good thing is that I fully know where I should add some subroutines to solve some minor problems. The bad thing is (extremely bad) that it took me two years to write it, I should evaluate it a lot first, and at the moment I haven't started simulating combustion (from March). So be happy that you have an important tool for your research.

Are you a phd-student or a researcher?


MohsinMukhtar January 7, 2010 06:37

Hello

I hope you people are doing well. Actually I m new to GAMBIT AND FLUENT. I needed to ask you about GAMBIT drawing.

1. My journal files are too messed up (like while making a drawing I undo several times so its all written in journal files) How can I omit and clean journal files so that whatever finally i did should be written over there. Also how I have made quite a complex drawing so how to search a particular work(like editing an arc etc) in Journal files.

2. I made a frustum. How to invert it.

3. How to de unite volumes.

4. I worked in Solid Edge there u can put dimensions but What about Gambit. Can we put dimensions on figures.

I will be grateful for your reply. Thank you very much
MOHSIN MUKHTAR


All times are GMT -4. The time now is 21:02.