Hello everyone

I am very new to grid generation. I am trying to use the cartesian grid generation method to generate a 2d mesh for a 2d object. However I cant seem to find any literature on the cartesian method. Could somebody please give me advice on any literature available, or how I might be able to approach the problem. Thanking u in advance.

Abu

(1). First, take a look at the Resources/grid generation section, there are many websites you can visit. (2). What is this " cartesian grid generation method " ?

i am trying to convert a y-shaped pipe from the physiacl domain to the computational domain. At the moment I have generated the boundary points for the pipe but now I am not sure how to generate the pipe in the computational domain. I have been told by my project supervisor that the cartesian method would be the easiest method to use to do this, however I have not found any literture on this method. Am i correct in assuming this method is a type of structured grid strategy. I would really appreciate any help. I have checked the cfd resources but unfortunately the NASA website on cartesian grids is not accessible

abu

(1). It is a good idea for you to check with your project supervisor to find out this method. I think, everyone has his own definition. It's hard for me to guess at it.

Unfortunately my project supervisor wants me to research into the method myself as I will get marked on this, so I would really appreciate any help. thanx

See the website for cfdrc. They have a way to generate cartesian grids. Even icem is coming up with a cartesian mesh generater.

(1). The Cartesian grid (if you are talking about using vertical and horizonatl Cartesian coordinates as grid lines) is not suitable for curved boundary, even if one can refine it near the wall. (2). This is because the viscous effect is of boundary layer type, which requires fine mesh, body-aligned boundary layer coordinates. So, for very low Reynolds number flow problems, it might be all right to use it. (3). But in general, you will have to use body-aligned coordinates. Even with the more flexible unstructured tet/tri mesh, it still requires hex/quad type next to the wall in order to make it work for high Reynolds number flows in the near the wall region.

Search for Marsha Berger in the literature she is one person I know who's done lots of research in that area. also CFDRC (http://www.cfdrc.com) has a cartesian grid generator