[Nek5000-users] New Recycling BC method?

nek5000-users at lists.mcs.anl.gov nek5000-users at lists.mcs.anl.gov
Sun Jun 20 20:12:14 CDT 2010


Hi Michael,

We're working on an interpolation-based approach.

>From the file you sent I can see that it could in general
be confusing w/ the old approach.

Paul

PS - was the file you sent just a test, or for production
      purposes?  I think more refinement near the walls would
      normally be in order.



On Sun, 20 Jun 2010, nek5000-users at lists.mcs.anl.gov wrote:

> Hi Paul,
>
> Were you able to find a solution for this?
>
> I was able to establish a new connectivity matrix for connecting an 
> arbitrary inlet plane to an outflow plane that you specifiy by a number of 
> elements. This could be called once at the beginning. I'm not sure how to 
> use it though but it is a positive start.
>
> - Michael
>
> On Jun 16, 2010, at 10:00 PM, nek5000-users at lists.mcs.anl.gov wrote:
>
>> Sounds great!  Thanks for looking into this!
>> 
>> 
>> 
>> - Michael
>> 
>> 
>> ----- Original Message -----
>> From: nek5000-users at lists.mcs.anl.gov
>> To: nek5000-users at lists.mcs.anl.gov
>> Sent: Wednesday, June 16, 2010 9:58:53 PM GMT -06:00 US/Canada Central
>> Subject: Re: [Nek5000-users] New Recycling BC method?
>> 
>> Michael, got it... I should be able to send a fix by morning that will 
>> work for any instances in which slab at location A has same (arbitrary) 
>> topology orientation as slab at location B, as is the case in the example 
>> you sent. This won't cover all cases but would likely cover 90%. hth, Paul 
>> On Wed, 16 Jun 2010, nek5000-users at lists.mcs.anl.gov wrote: > > > Paul, > 
>> > > > See attached for the rea. Thanks, > > > > Michael > > > ----- 
>> Original Message ----- > From: nek5000-users at lists.mcs.anl.gov > To: 
>> nek5000-users at lists.mcs.anl.gov > Sent: Wednesday, June 16, 2010 9:47:29 
>> PM GMT -06:00 US/Canada Central > Subject: Re: [Nek5000-users] New 
>> Recycling BC method? > > 960 is ok On Wed, 16 Jun 2010, 
>> nek5000-users at lists.mcs.anl.gov wrote: > > > Paul, > > > > The rea I have 
>> now from matlab/gambit has 960 elements.  Do you want me to create 
>> another, smaller rea that may make it easier to work with? > > > > The 
>> geometry by the way is a simple pipe flow. > > > > - Michael > > > ----- 
>> Original Message ----- > From: nek5000-users at lists.mcs.anl.gov > To: 
>> nek5000-users at lists.mcs.anl.gov > Sent: Wednesday, June 16, 2010 9:39:59 
>> PM GMT -06:00 US/Canada Central > Subject: Re: [Nek5000-users] New 
>> Recycling BC method? > > Hi Micheal, I agree completely that it's best to 
>> avoid mix & match meshing. If you want to send me a gzipped .rea file 
>> produced by Gambit, I can take a look and see if I have any ideas. Paul On 
>> Wed, 16 Jun 2010, nek5000-users at lists.mcs.anl.gov wrote: > > > Hi Paul, > 
>> > > > Thanks for the reply.  The reason for the interest is due to the 
>> Gambit to Nek matlab code. We have had great success importing complex 
>> geometries with it, but no w we run into the issue of the recycling 
>> boundary condition.  Gambit's ordering of the elements is not always 
>> consistent which makes it impossible to use the current technique. > > > > 
>> Our solution was to create the portion that has the recycling in prex and 
>> then merge it with the complex part created in Gambit. However, this is 
>> proving to be very time consuming and troublesome (Element Mismatch 
>> Errors).  So, to avoid this I would like to come up with a new recycling 
>> method that does not depend on the elements being in a specific order. My 
>> u nderstanding is that the current method looks at the inflow, counts a 
>> certain number of elements up, copies this outflow from face 6 and puts it 
>> back into the inflow face 5. > > > > What I would like to do is replace 
>> how the inflow plane is connected to the outflow plane. My idea was to use 
>> the connectivity matrix to figure out what the element numbers and outflow 
>> faces are at a certain distance away from the inflow plane, and use t hese 
>> faces to connect.  The cross-section for the inflow and outflow plane 
>> would be the same, but you couldn't necessarily say that it is located, 
>> say inflow + 30 elements for example, OR that the outflow and inflow faces 
>> are all the same (face 5 for inflow, face 6 for outflow for example).  > > 
>> > > Do you think this is possible? In a way, you would be creating a new 
>> connectivity matrix that governs the recycling BC. > > > > Thanks! > > - 
>> Michael M. > > > ----- Original Message ----- > From: nek5000-users at lists. 
>> mcs . anl .gov > To: nek5000-use rs at lists. mcs . anl .gov > Sent: 
>> Wednesday, June 16, 2010 9:08:00 PM GMT -06:00 US/Canada Central > 
>> Subject: Re: [Nek5000-users] New Recycling BC method? > > Hi Michael, 
>> Access to the connectivity array wouldn't help in this case, given that 
>> the recycling bc needs to establish a _new_ connectivity that has nothing 
>> to do with the existing one. The examples use this approach of 
>> establishing a new connectivity. I'm confident that we ca n make the 
>> existing approach for any mesh that can legitimally employ the recycling 
>> bc - i.e., any mesh that has the same cross-sectional distribution of 
>> elements at the two planes of interest. An alternative is to use the 
>> general purpose interpolation routines. There may (or may not) be a 
>> performance issue -- we're looking into this w/ the help of a summer 
>> student. Please let me know if I can assist you w/ your specific case. 
>> Paul > I am still trying to figure this idea out, but I think I would need 
>> to > access the connectivity/boundary condition information that is 
>> contained > in the rea file. Does anyone know the array, or what to call 
>> in order to > access this in the usr file?  Thanks again for a ny help! > 
>> > > > - Michael M. > > > ----- Original Message ----- > From: 
>> nek5000-users at lists. mcs . anl .gov > To: " Nekton User List" > Sent: 
>> Tuesday, June 15, 2010 9:02:00 PM GMT -06:00 US/Canada Central > Subject: 
>> [Nek5000-users] New Recycling BC method? > > > > > Hello All, > > > > I am 
>> wondering if there is another technique to set up a recycling boundary 
>> condition that does not depend on the elements being in a certain order 
>> (say based on element coordinates for instance)?   I would like to pick 
>> the recycling plane based on a given offset distance rather than a number 
>> of elements. > > > > Also, the outflow plane faces may not all necessarily 
>> be face 6 (and face 5 for the inflow plane). > > > > I have tried going 
>> through the examples where a recycling BC is used however I am finding it 
>> difficult to understand as my knowledge of Fortan is still growing. Thanks 
>> for any help anyone can provide with this idea. > > > > Regards, > > 
>> Michael M. > > >   > ___________________ ____________________________ 
>> Nek5000-users mailing list Nek5000-users at lists. mcs . anl .gov https 
>> ://lists. mcs . anl .gov/mailman/ listinfo /nek5000-users > 
>> _______________________________________________ Nek5000-users mailing list 
>> Nek5000-users at lists. mcs . anl .gov https :/ /lists. mcs . anl 
>> .gov/mailman/ listinfo /nek5000-users > 
>> _______________________________________________ Nek5000-users mailing list 
>> Nek5000-users at lists.mcs.anl.gov 
>> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > 
>> _______________________________________________ Nek5000-users mailing list 
>> Nek5000-users at lists.mcs.anl.gov 
>> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users
>> _______________________________________________ Nek5000-users mailing list 
>> Nek5000-users at lists.mcs.anl.gov 
>> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users
>> 
>> _______________________________________________
>> Nek5000-users mailing list
>> Nek5000-users at lists.mcs.anl.gov
>> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users



More information about the Nek5000-users mailing list