extrude by providing z-coordinates

[Weixiong Zheng]

Hi,

I found out there is not a function for extruding triangulation by exact position of cell vertices along z-axis. So I wrote a piece of code by modifying

GridGenerator::extrude_triangulation.

The attachment is an example of what it does.

Did for my own needs. Just wonder if I can contribute to deal.II with this super trivial work. Never mind if it's too trivial to put it in deal.II or someone else if already doing similar work.

Best,

Weixiong

[Wolfgang Bangerth]

Any contribution is much appreciated!

If I understand correctly, you just generalize the second and third
arguments of the existing function

void
extrude_triangulation (const Triangulation<2, 2> &input,
const unsigned int n_slices,
const double height,
Triangulation<3,3> &result);

? I can totally see how that makes sense!

Best
W.


--
------------------------------------------------------------------------
Wolfgang Bangerth email: bang...@colostate.edu
www: http://www.math.colostate.edu/~bangerth/

[Weixiong Zheng]

Sir,

Yes. It then allows non-uniform slicing along z-axis.

Also, I am wondering if you guys are interested in making triangulation based on a basic one for repeated structure as shown in the attachment. I have to do this for my project but didn't find a function so I did something like copy->shifting->merging->extruding (if 3D). I think it would be cool to have this in deal.II.

You have any instruction on how to make contribution?

Weixiong

在 2018年4月4日星期三 UTC-7下午1:47:37，Wolfgang Bangerth写道：

[Wolfgang Bangerth]

Weixiong,

> Yes. It then allows non-uniform slicing along z-axis.

Great. Yes, please do submit a pull request for this function! (And let
us know if you need help figuring out how to create pull requests! I
know we have a how-to file somewhere, but I forgot where exactly it is.
Someone?)

> Also, I am wondering if you guys are interested in making triangulation
> based on a basic one for repeated structure as shown in the attachment.
> I have to do this for my project but didn't find a function so I did
> something like copy->shifting->merging->extruding (if 3D). I think it
> would be cool to have this in deal.II.

Yes. I had to laugh at your previous picture which shows that you're
still working with fuel assemblies apparently -- and I wanted to say
"no, no, you'll need 17x17 replications, not 5x5!"

I think it would be neat to have a function of the form replicate(...)
that does exactly what you are suggesting and are already doing in your
code. It would take an existing triangulation and replicate it a number
of times in the x, y, and z-directions.

[Weixiong Zheng]

Sir

The Issue 6158 and 6159 have been opened for the proposed functionalities. Tell you if you have any

thought.

By instruction, it's actually more for the testing. Maybe I will just follow how the other functions are tested.

By the way, I am at UC Berkeley leading an open source group project based on deal.II on GitHub 

called "Bay-Area Radiation Transport (BART)" with MIT license for research/mentoring purpose. The link

is where the repo is "https://github.com/SlaybaughLab/BART". You may add it to where you tell which 

projects are using deal.II if you want.

Weixiong

在 2018年4月4日星期三 UTC-7下午2:55:44，Wolfgang Bangerth写道：

[Wolfgang Bangerth]

Weixiong,

> By instruction, it's actually more for the testing. Maybe I will just
> follow how the other functions are tested.

Yes -- my took of choice is generally to do
grep -r merge_triangulations tests/*/*cc
to find out which tests use a given function (here,
merge_triangulations()). I then typically clone such a test and modify
it to use the function I'm currently writing.

> By the way, I am at UC Berkeley leading an open source group project
> based on deal.II on GitHub
> called "Bay-Area Radiation Transport (BART)" with MIT license for
> research/mentoring purpose. The link
> is where the repo is "https://github.com/SlaybaughLab/BART". You may add
> it to where you tell which
> projects are using deal.II if you want.

Excellent -- I've done this now!

Thanks

[Weixiong Zheng]

PR has been approved and will be merged sometime. Thanks for fill

the project name into Applications entry.

在 2018年4月5日星期四 UTC-7下午3:48:05，Wolfgang Bangerth写道：