[Nek5000-users] Pipe with expansion

nek5000-users at lists.mcs.anl.gov nek5000-users at lists.mcs.anl.gov
Sun Jan 19 22:41:49 CST 2014


Kamal,

Looks good.

I'm guessing that one further refinement would be:

if (z.gt.z0) scale = 1 + (smax-1)*(z-z0)/(z1-z0)
if (z.gt.z1) scale = smax

which would yield

    scale = 1 when z=z0
    scale = smax when z >= z1

Paul



On Sun, 19 Jan 2014, nek5000-users at lists.mcs.anl.gov wrote:

> Hi paul,
>
> The code works fine, but there should be a small change thats has to be done
>
>
> if (z.gt.z0) scale = 1 + (z /(z1-z0 )))
> if (z.gt.z1) scale = smax
>
> This creates a perfect diverging pipe.
>
> Thanks for the help paul.
>
> Kamal
>
> On Jan 20, 2014, at 12:09 AM, nek5000-users at lists.mcs.anl.gov wrote:
>
>> Hi paul,
>>
>> If I am not wrong,
>> I make a mesh with three straight pipes( inlet, divergent, outlet ) connected together via n2to3 so that I know the element connectivity places of the divergent section (z0,z1) and use the section what you gave
>>
>> Thank you for the code
>>
>> Kamal
>>
>> On Jan 19, 2014, at 8:45 PM, nek5000-users at lists.mcs.anl.gov wrote:
>>
>>>
>>> Kamal,
>>>
>>> I would do the following in usrdat2():
>>>
>>>     n=nx1*ny1*nz1*nelt
>>>
>>>     do i=1,n
>>>        x=xm1(i,1,1,1)
>>>        y=ym1(i,1,1,1)
>>>        z=ym1(i,1,1,1)
>>>        scale = 1.
>>>        if (z.gt.z0) scale = smax*(z-z0/(z1-z0)
>>>        if (z.gt.z1) scale = smax
>>>        xm1(i,1,1,1) = scale*x
>>>        ym1(i,1,1,1) = scale*y
>>>     enddo
>>>
>>> Paul
>>>
>>> Make certain that z1 and z0 correspond to z values
>>> that precisely coincide with element interfaces.
>>>
>>>
>>> On Sun, 19 Jan 2014, nek5000-users at lists.mcs.anl.gov wrote:
>>>
>>>> Hi Paul,
>>>>
>>>> I looked into the Peris example but It is so what same like the stenosis example with a Cosine deformation.
>>>>
>>>> In my case it is a pipe with conical divergent. I tried to play around the variables Xm1, Ym1 and Zm1 but the mapping doesn't go well.
>>>>
>>>> Could you please help me with this.
>>>>
>>>> Thank you,
>>>>
>>>> Kamal
>>>>
>>>>
>>>> On Jan 18, 2014, at 7:26 PM, nek5000-users at lists.mcs.anl.gov wrote:
>>>>
>>>>> Hi Paul,
>>>>>
>>>>> I looked at the Helix and stenosis example but they are using cosine functions to deform
>>>>>
>>>>> I tried to use a tangent function but it bends the whole pipe.
>>>>>
>>>>> I dint look at peris example .
>>>>>
>>>>> I will take a look at it and let you know.
>>>>>
>>>>> Thank you
>>>>>
>>>>> Kamal
>>>>>
>>>>> On Jan 18, 2014, at 7:22 PM, nek5000-users at lists.mcs.anl.gov wrote:
>>>>>
>>>>>>
>>>>>> Kamal,
>>>>>>
>>>>>> Have you looked at the peris example?  The usr file there shows how to deform
>>>>>> a pipe geometry.
>>>>>>
>>>>>> Paul
>>>>>>
>>>>>> From: nek5000-users-bounces at lists.mcs.anl.gov [nek5000-users-bounces at lists.mcs.anl.gov] on behalf of nek5000-users at lists.mcs.anl.gov [nek5000-users at lists.mcs.anl.gov]
>>>>>> Sent: Saturday, January 18, 2014 12:17 PM
>>>>>> To: nek5000-users at lists.mcs.anl.gov
>>>>>> Subject: Re: [Nek5000-users] Pipe with expansion
>>>>>>
>>>>>> Dear Paul,
>>>>>>
>>>>>> I tried many ways to map my straight pipe geometry but it does not seem to work properly
>>>>>>
>>>>>> I used something like
>>>>>>
>>>>>> X = X + Z * tan ( alpha )
>>>>>> Y = Y + Z* tan (alpha )
>>>>>>
>>>>>>
>>>>>> even that dint work. could please help me with this
>>>>>>
>>>>>> Thank you
>>>>>>
>>>>>> Kamal.
>>>>>>
>>>>>> On Jan 17, 2014, at 2:24 PM, nek5000-users at lists.mcs.anl.gov wrote:
>>>>>>
>>>>>>> Hi Paul,
>>>>>>>
>>>>>>> Thanks paul. I tried working with the diverging mesh modification.
>>>>>>>
>>>>>>> I tried to use this form of equation to modify the mesh
>>>>>>>
>>>>>>> z = mx + b
>>>>>>> z = ny  + c
>>>>>>>
>>>>>>> x = ( z - b ) / m
>>>>>>> y = (z - c)  / n
>>>>>>>
>>>>>>> Do you think would work ? or is there any other way to input that ?
>>>>>>>
>>>>>>> Thanks,
>>>>>>>
>>>>>>> Kamal
>>>>>>>
>>>>>>>
>>>>>>> On 17/01/2014 14:14, nek5000-users at lists.mcs.anl.govwrote:
>>>>>>>> Kamal,
>>>>>>>>
>>>>>>>> 1) To check for convergence, run your simulation for increasing polynomial orders ( lx1 in the SIZE file) and measure some consistent quantity (drag, TKE, etc).  You can plot this quantity against lx1 and this should give you a good picture of how you are converging (without needing to change the physical mesh).
>>>>>>>>
>>>>>>>> 2) If you weren't using temperature in your simulation, you could do put the following code snippet in your userchk
>>>>>>>>
>>>>>>>> c-------------------------------------------
>>>>>>>> integer i, ntot
>>>>>>>>
>>>>>>>> ntot = nx1*ny1*nz1*nelv  ! total number of points in simulation (on THIS processor)
>>>>>>>>
>>>>>>>> do i = 1,ntot
>>>>>>>>  t(i,1,1,1,1) = 0.5*(vx(i,1,1,1)**2 + vy(i,1,1,1)**2 + vz(i,1,1,1)**2)
>>>>>>>> enddo
>>>>>>>>
>>>>>>>> ifto = .true.
>>>>>>>> c-------------------------------------------
>>>>>>>>
>>>>>>>> I've assumed here that your density is unity.  This computed kinetic energy point-wise and stores it in the temperature field.  The part "ifto = .true." tells Nek to output the temperature field whenever it outputs a field file.  You could then view this in visit just like you would the velocity field.
>>>>>>>>
>>>>>>>> Hope this helps!
>>>>>>>>
>>>>>>>> Josh
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> On Wed, Jan 15, 2014 at 9:52 AM, <nek5000-users at lists.mcs.anl.gov> wrote:
>>>>>>>> Hi Paul,
>>>>>>>>
>>>>>>>> I will try that mapping, I created a straight pipe to do some simulations.
>>>>>>>>
>>>>>>>> 1) Is there a way to check my spectral convergence, so that I can do some mesh convergence study?
>>>>>>>> 2) Is there a way to plot the kinetic energy in the flow ?
>>>>>>>>
>>>>>>>> Thank you,
>>>>>>>> Kamal
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> On 15/01/2014 12:14, nek5000-users at lists.mcs.anl.gov wrote:
>>>>>>>>
>>>>>>>> Hi Kamal,
>>>>>>>>
>>>>>>>> For the geometry you describe, I would apply mesh morphing
>>>>>>>> to a standard straight pipe.  You can create such a pipe
>>>>>>>> using a subset of the mkmesh script provided in the expansion
>>>>>>>> example.
>>>>>>>>
>>>>>>>> If your transition region is conical, then I woud make
>>>>>>>> certain that the change in slope occurs at element boundaries,
>>>>>>>> i.e., as in the half cross-section depicted below:
>>>>>>>>
>>>>>>>>                             _______________
>>>>>>>>                            /|      |      |
>>>>>>>>                          /  |      |      |
>>>>>>>>                        /|   |      |      |
>>>>>>>>      ________________/  |   |      |      |
>>>>>>>>      |     |     |   |  |   |      |      |
>>>>>>>>      |     |     |   |  |   |      |      |
>>>>>>>>      +-----+-----+---+--+---+------+------+---->  z
>>>>>>>>
>>>>>>>>
>>>>>>>> Paul
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> On Wed, 15 Jan 2014, nek5000-users at lists.mcs.anl.gov wrote:
>>>>>>>>
>>>>>>>> Hi Paul,
>>>>>>>>
>>>>>>>> I looked at the example and it looks very helpful, but for a sudden expansion we have only two parts (i.e) the inlet pipe and the outlet pipe. Whereas in a diverging pipe we have three section (i.e,) The inlet pipe, Diverging section and the outlet pipe.
>>>>>>>>
>>>>>>>>
>>>>>>>> Is it possible for me to create a 2D diverging section using mesh modification and revolve it around the centre axis to get a 3d diverging pipe ?
>>>>>>>>
>>>>>>>>
>>>>>>>> or should I apply the mesh modification directly to the 3d sudden expansion pipe using usrdat2() ?
>>>>>>>>
>>>>>>>> Thanks,
>>>>>>>> Kamal
>>>>>>>>
>>>>>>>>
>>>>>>>> On 14/01/2014 21:09, nek5000-users at lists.mcs.anl.gov wrote:
>>>>>>>> Hi Paul,
>>>>>>>>
>>>>>>>> Thanks a lot for the update. I will take a look at as soon as possible and keep you updated with it.
>>>>>>>>
>>>>>>>> Thank you,
>>>>>>>>
>>>>>>>> Kamal
>>>>>>>>
>>>>>>>> On Jan 14, 2014, at 9:04 PM, nek5000-users at lists.mcs.anl.gov wrote:
>>>>>>>>
>>>>>>>> Kamal,
>>>>>>>>
>>>>>>>> I've added an example for a step-type expansion in a pipe. New example directory
>>>>>>>> is nek5_svn/examples/expansion, with a README.
>>>>>>>>
>>>>>>>> If you want a standard diffuser, I would use mesh-morphing, as outlined in the
>>>>>>>> nek primer, which can be found on this page:
>>>>>>>>
>>>>>>>> http://nek5000.mcs.anl.gov/index.php/UG#Documentation
>>>>>>>>
>>>>>>>> I've also automated our turbulent outflow bc treatment -- one simply provides an
>>>>>>>> expansion factor (typically ~ 1.5) and two persistent arrays to turb_outflow(), which
>>>>>>>> is now in navier5.f.    This can be useful when very strong vortices are leaving the
>>>>>>>> domain, which is possible for certain diffuser flow conditions (and others).
>>>>>>>>
>>>>>>>> Paul
>>>>>>>> ________________________________________
>>>>>>>> From: nek5000-users-bounces at lists.mcs.anl.gov [nek5000-users-bounces at lists.mcs.anl.gov] on behalf ofnek5000-users at lists.mcs.anl.gov [nek5000-users at lists.mcs.anl.gov]
>>>>>>>> Sent: Monday, January 13, 2014 3:35 AM
>>>>>>>> To: nek5000-users at lists.mcs.anl.gov
>>>>>>>> Subject: [Nek5000-users] Creating a Mesh
>>>>>>>>
>>>>>>>> Hi
>>>>>>>>
>>>>>>>> Can some one please help me in this because I am new to nek
>>>>>>>>
>>>>>>>> 1) How can I create a .rea for a specific geometry say ' A sudden
>>>>>>>> expansion pipe ' or a ' Diverging pipe '
>>>>>>>>
>>>>>>>> 2) What tool I should use to create that mesh.
>>>>>>>>
>>>>>>>>
>>>>>>>> Thank you
>>>>>>>>
>>>>>>>>
>>>>>>>> kamal
>>>>>>>> _______________________________________________
>>>>>>>> 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
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> --
>>>>>>>> Josh Camp
>>>>>>>>
>>>>>>>> "All that is necessary for the triumph of evil is that good men do nothing" -- Edmund Burke
>>>>>>>>
>>>>>>>>
>>>>>>>> _______________________________________________
>>>>>>>> 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
>
> _______________________________________________
> 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