[overture] Re: Two pipes
- From: Bill Henshaw <wdhenshaw@xxxxxxxxx>
- To: overture@xxxxxxxxxxxxx
- Date: Fri, 10 Oct 2014 17:27:52 -0400
Hi Stone,
It would be very complicated to make the disk actually touch the boundary
for only part of the simulation -- if you make the gap small then there
should be little flow through the gap. The gap can get smaller as you refine
the mesh.
I suppose you could add an additional drag force in the gap when it is
closed
to further reduce the flow. By changing the gap width you could evaluate
how accurate your simulation is. Remember that there are many approximations
being made in your model already so a thin gap is just one more
approximation.
Regards,
Bill
On Fri, Oct 10, 2014 at 3:58 PM, Shi Qiu <shiq@xxxxxxx> wrote:
> Hi Bill,
>
> Is there a way to make the disk surface touch the cylinder? Because I want
> to simulate a switch (disk in the pipe). When it closes, there is no flow.
> When it opens with an angle, I want to find out the pressure distribution
> on it. In the latter situation, some parts of the disk will tough the
> cylinder. Any suggestions?
>
> Thanks a lot.
>
> Best,
>
> Stone
>
> On Fri, Oct 10, 2014 at 6:47 AM, Bill Henshaw <wdhenshaw@xxxxxxxxx> wrote:
>
>> Hi Stone
>> It sounds like you have figured out how to build a disk -- see
>> also pillInABoxGrid.cmd for a grid for a disk.
>> You will need to decrease the diameter of the disk surface so
>> that it does not touch the cylinder -- so there is a gap at of least 3-5
>> grid lines and
>> then you should be able to construct the grid. It doesn't
>> matter if the disk volume grid goes outside the cylinder (only the
>> disk surface needs to be inside) as ogen will cut away this part.
>>
>> Regards
>> Bill
>>
>>
>>
>> On Fri, Oct 10, 2014 at 1:51 AM, Shi Qiu <shiq@xxxxxxx> wrote:
>>
>>> Hi Bill and Mike,
>>>
>>> I am sure that you are really busy and I appreciate you taking the time
>>> to respond personally to me.
>>>
>>> I have been studying these samples for a week. Now I have some ideas
>>> about how to generate the grid I want. Originally the goal is to generate a
>>> pipe and put a solid disk in the pipe with an angle, the diameter of the
>>> disk is same with pipe (which means some parts of the disk are tangent with
>>> the pipe). I want to use cgins to find out pressure distribution on the
>>> disk when there are uniform flow in the pipe. The first approach I tried
>>> was to use one thin pipe as disk and set its boundaries as physical
>>> boundary, then I put it in another pipe. But this does not work. Then I
>>> found the body of revolution function. Now I am trying to use this function
>>> to build a dick (similar to revolve.cmd in sampleGrids folder). Then use
>>> rotate/scale/shift function to rotate a pipe and put the disk in it.
>>> However, I have some problems when I want to put the dick in the pipe with
>>> the boundaries tangent to the pipe. This is because of there are body
>>> fitted grids around the dick. When it is tangle to the pipe, these grids
>>> will be outside the pipe. Could you give me some suggestions please? Is
>>> there a better approach to generate these grids?
>>>
>>> Thanks again,
>>>
>>> Best,
>>>
>>> Stone
>>>
>>> On Wed, Oct 1, 2014 at 9:14 AM, Mike Singer <msinger2006@xxxxxxxxx>
>>> wrote:
>>>
>>>> Hi Stone,
>>>>
>>>> I remember doing something similar a few years ago. I just put the
>>>> pipes close to each other (and overlapping a bit), and ogen figured out the
>>>> hole cutting and made a nice overlapping grid. You might want to give it a
>>>> try since having ogen figure it out is probably the most straightforward
>>>> approach. Btw, you'll likely have to play with the amount of overlap, grid
>>>> resolution, etc. in order to make it work.
>>>>
>>>> Mike
>>>>
>>>>
>>>>
>>>> On Wed, Oct 1, 2014 at 7:39 AM, Bill Henshaw <wdhenshaw@xxxxxxxxx>
>>>> wrote:
>>>>
>>>>> Hi Stone,
>>>>>
>>>>> twoPipes.cmd - uses a special analytically defined fillet grid,
>>>>> might be possible to
>>>>> work out the formula or pipes at an angle.
>>>>>
>>>>> filletTwoCyl.cmd - uses the more generic FilletMapping which may
>>>>> work if the pipes are at an angle,
>>>>> this example creates grids "outside" the pipes but
>>>>> this can be changed to do inside.
>>>>>
>>>>> joinTwoCyl .cmd - another example with a sharp corner join,
>>>>> if you want the region exterior to the two pipes.
>>>>>
>>>>>
>>>>> If you can get a IGES (CAD) file of the geometry then you can
>>>>> create a grid with the
>>>>> hyperbolic grid generator.
>>>>>
>>>>> ...Bill
>>>>>
>>>>> On Tue, Sep 30, 2014 at 2:22 AM, Shi Qiu <shiq@xxxxxxx> wrote:
>>>>>
>>>>>> Hi Bill,
>>>>>>
>>>>>> In the Overture sample grids folder. I saw the sample of two pipes
>>>>>> intersect with each other (3D). I wonder that if we can change the angle
>>>>>> of
>>>>>> two pipes, make them not perpendicular with each other. Could you give me
>>>>>> some suggestions about this problem please?
>>>>>>
>>>>>> Best,
>>>>>>
>>>>>> Stone
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>
>>>>
>>>
>>
>
Other related posts:
