[dengdeng]

Hi,

I am simulating a case with a deforming plate in Fluent. The deformation of the boundary is realized by a UDF of DEFINE_GEOM, which is already built and compiled successfully.

At the very beginning, I worked with the following mesh:


The UDF is proved to be OK as the Preview Mesh Motion icon being clicked. However, as the zoomed-in view shown on the right side of this image, the mesh edge adjacent to the trailing edge point of the deforming plate also inclined with it, which is originally expected to be unmoved. (The inclined edge is marked with yellow line, and the plate with white line. The red dashed line represents the original position of the mesh. The same problem also happened to the leading edge region)

To avoid this unexpected motion of the adjacent grid, I improved the structure of my mesh by introducing interfaces like this:


Well, this time, a rather serious problem occurs that the mesh disappears as Preview Mesh Motion executed:

and there's no use to display all surfaces through Mesh display.

Has anyone met with similar problems before? Any suggestion is appreciated.

Thank you.

[dengdeng]

It seems I failed to upload the pictures. Anyway, the images mentioned are:

https://wx4.sinaimg.cn/large/9c7300f...20n10fd76g.jpg

https://wx3.sinaimg.cn/large/9c7300f...20np0eu77k.jpg

https://wx2.sinaimg.cn/large/9c7300f...20kz0guq3f.jpg

Please let me know if you can not acess the pictures, thanks.

[dengdeng]

Well, the problem has been solved by changing the expression of my UDF. Originally, I defined the variable time like this:

real time = CURRENT_TIME;

Then my colleague helped to redefine it as:

double t =  RP_Get_Real("flow-time");

I do not know the exact reason though, this modification does work well.

[blackmask]

You have changed the variable from "time" to "t", which makes the difference. The macro expansion for CURRENT_TIME is nothing but "RP_Get_Double("flow-time")".