Regarding mesh movemsnt in I.C Engine cylinder
 

Dear CFX user, I am working with a project on Piston & valve movement for 4 stroke IC engine using CFX-5.6. I am trying to make the piston and valve movemnt by generating codes. Though I able to get the piston and valve movement but not at at desired level. Reason is the mesh elements provide negative volume and get crashed after covering a distance up to about 30 % of cylinder height. Could any one CFX user guide me the right way to sort out the mentioned problem.Whether i should opt for mesh elements like Hexa with unstructured or so. Please mail me with proper solution on amishra2k@yahoo.co.in Thanks & Regards Ashwini Mishra

Re: Regarding mesh movemsnt in I.C Engine cylinder
 

Hi Ashini,

The moving mesh feature of CFX5.6 is a beta feature only so it is not perfect. It will be a fully supported feature in 5.7 when it comes out, however.

I found to mesh large mesh deformations (like that found in IC engines), you have to draw your initial mesh at top dead centre. This then means you have to draw a mesh which still gives a decent mesh when stretched out to bottom dead centre. You can't do this with tet elements, so you need to mesh across the cylinder plane with either tris or quads, then extrude that mesh down the cylinder axis with a very fine mesh. This makes the mesh at TDC have a very large aspect ratio, but this is unavoidable without cell addition/deletion.

It's a tricky meshing operation, but it is the only way of doing it. I have considered interpolating to a different mesh half way through the stroke, but have not tried it yet and it will just add to the already significant degree of difficulty.

Good luck! Glenn

Re: Regarding mesh movemsnt in I.C Engine cylinder
 

Hi there,


Glenn is right.


I worked once generating only one grid and moving it along time (or crank angle, whatever) but I spent a huge time to find the optimal mesh to make myself able to finish one entire engine cycle. Another way is to make several grids, as Glenn also stated, each grid to its specific position and interpolate them as long as you proceed with the solution. However, this could be really a pain in the a** to do it in a HEXA mesh. I saw some time ago someone who made scripts automatizing this operation in ICEM Tetra. But, of course, I'm sure he also spent a lot of time working on this automatization task.
Take care, Gloria Gaynor

Re: Regarding mesh movemsnt in I.C Engine cylinder
 

You can re-invent the wheel and use cfx or you can just use starcd which is used by most automotive companies to run this type of calculation. There are some good articles about this topic on www.adapco-online.com

Re: Regarding mesh movemsnt in I.C Engine cylinder
 

Glenn, quick question as I don't understand the geometry. Do you have to deform the whole mesh to generate the piston cyle, or could you translate the mesh around the piston, deforming the mesh in areas away from the the piston (assuming the piston is where the majority of the mesh would be ?). Just wondering if this would make like any easier ? when we did some moving mesh work around a complex shape, that was the approach we had to take. Bob