Modeling movement of compressor valve

Marco Müller

Hi,

i'm trying to model the movement of compressor valves by CEL with mesh deformation. I use one domain. The valve (see picture) is defined as boundary with specified displacement (CEL). I had no idea how to model the elements for the flow when the valve is openened, so I tried to keep it already open to the beginning of the simulation, and to "expand" the elements in the crack (beginning height 0.2 mm). (expanding factor 25. The beginning aspect ratio of the elements in the crack is 1:5, in the end 5:1) However the end is never reached, because the poor elements surrounding the crack get negative volumes (yet if the valve is opened 1 mm) and solution stops. My question is now, whether there exists a better way to model this problem ("creating elements ex nihilo") or how to realize it with the existing method. Thank You for your replies!!!

http://img355.imageshack.us/my.php?image=01zd7.jpg

http://img604.imageshack.us/content..../179/01zd7.jpg

http://img100.imageshack.us/my.php?image=02iu4.jpg

http://img604.imageshack.us/content....5892/02iu4.jpg

http://img100.imageshack.us/my.php?image=03by8.jpg

http://img604.imageshack.us/content..../953/03by8.jpg