[Nek5000-users] Add a forcing term on scalar equation

Mon Jan 23 04:15:06 CST 2012

Dear Stefan,

We are solving a system like this:

Dc/Dt= Nabla^2 c +NL(u,c)

DU/Dt= Nabla^2 u - grad p + f(c), where we want to use the c just computed.

It's okay to solve the scalar equation explicitly according to previous tests with another fully spectral (Fourier) code. But, it's difficult to get stable results if the momentum equation is fully explicit.

Two questions: would it be a problem to use Pn-Pn? Would it be slower or less accurate?

In the code, can we change the forcing term to be implicit, so using Crank-Nicolson for f(c) (f contains derivatives).

We understood userf is adding terms to the AB part, while we want something implicit.

Cheers
Iman

________________________________________
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: Friday, January 20, 2012 5:24 PM
To: nek5000-users at lists.mcs.anl.gov
Subject: Re: [Nek5000-users] Add a forcing term on scalar equation

Hi Iman,

I don't know what type of system you are trying solve but I want to
draw your attention on the following option:

You can use the PnPn formulation. First, a non-linear coupled scalar
system is solved using CVODE (a stiffly stable integrator).Then, the
hydrodynamic equations (velocity and pressure) are solved. The two
system are decoupled using a high-order splitting approach. We do the
same for chemically reactive flows.

Cheers,
Stefan

On 1/20/12, nek5000-users at lists.mcs.anl.gov
<nek5000-users at lists.mcs.anl.gov> wrote:
>
> Hi Iman,
>
> There are some minor bookkeeping issues here...
>
> Just to make certain we're on the same page, are you currently
> using Pn-Pn-2, or Pn-Pn ?
>
> Paul
>
>
> On Fri, 20 Jan 2012, nek5000-users at lists.mcs.anl.gov wrote:
>
>> Paul,
>>
>> Numerically the momentum equation has an explicit forcing term.
>> Only that, it is based on the T at the same time. However, I would like to
>> first solve the equation for the scalar.
>>
>> In other words, I want to first solve the equation for T, use it to
>> calculate f(T) and afterward solve the momentum equation ( u_t=NS + f(T)
>> ). So How I can do that!?
>>
>>
>> Iman
>>
>>
>>
>>
>>
>> ________________________________________
>> 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: Friday, January 20, 2012 2:25 PM
>> To: nek5000-users at lists.mcs.anl.gov
>> Subject: Re: [Nek5000-users] Add a forcing term on scalar equation
>>
>> Iman,
>>
>> These can be coupled explicitly, but not implicitly.   Usually, however,
>> this is sufficent.  Equation are of the form:
>>
>>    u_t = NS +  f(u,T)
>>
>>    T_t = energy eq. + q(u,T)
>>
>> f & q explicit.   The code will take care of the bookkeeping.
>>
>> Paul
>>
>>
>> On Fri, 20 Jan 2012, nek5000-users at lists.mcs.anl.gov wrote:
>>
>>> Thanks.
>>> So I just used the "USERQ" to implement the forcing term to scalar filed
>>> explicitly.
>>>
>>> In the next step, I want to coupled the scalar field to Navier-Stokes
>>> equation using a forcing term.
>>> In this case, the forcing term is a function of scalar field and I would
>>> like to implement it implicitly.
>>> As far as I know, if I used USERF, the scheme is explicit. So how can I
>>> add a forcing term to N-S equation implicitly!?
>>>
>>> Iman
>>>
>>>
>>>
>>> ________________________________________
>>> 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: Thursday, January 19, 2012 4:20 PM
>>> To: nek5000-users at lists.mcs.anl.gov
>>> Subject: Re: [Nek5000-users] Add a forcing term on scalar equation
>>>
>>> It's actually evaluated explicitly, so you shold be ok.
>>>
>>> Paul
>>>
>>>
>>> On Thu, 19 Jan 2012, nek5000-users at lists.mcs.anl.gov wrote:
>>>
>>>> Hi
>>>>
>>>> I would like to add a forcing term to the scalar equation.
>>>> According to the Nekton manual, Chapter 5, the forcing term is treated
>>>> implicitly and the convective term is integrated using the third order
>>>> Adam_Bashforth scheme (AB3).
>>>> For my case, the forcing term is a nonlinear function of the scalar
>>>> field, so I cannot use the implicit scheme and I should lump it with the
>>>> AB3.
>>>> Therefore, could you please guide me and indicate where in the code I
>>>> can find the implementation of AB3 for the convective term of passive
>>>> scalar?
>>>>
>>>> Cheers
>>>> Iman
>>>>
>>>>
>>> _______________________________________________
>>> 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