[openfoammaofnepo]

Hi All,

I met a problem when I read the EulerDdtScheme.C

https://github.com/OpenFOAM/OpenFOAM-2.2.x/blob/master/src/finiteVolume/finiteVolume/ddtSchemes/EulerDdtScheme/EulerDdtScheme.C

I found that when we do the time discretization, we need the rho and vf for both the current time step and the previous time step (lines 253 in the above wenpage):

rDeltaT*(rho*vf - rho.oldTime()*vf.oldTime())

So my question is: when OpenFOAm is running, when are rho.oldTime() and vf.oldTime() stored into these two fields as the old values? If we take the rhoPimpleFOAM as an example.

I checked the source files for GeometricField.C, and found that there are indeed the functions for storeOldTimes:

// Store old-time field
template<class Type, template<class> class PatchField, class GeoMesh>
void Foam::GeometricField<Type, PatchField, GeoMesh>::storeOldTimes() const
{
    if
    (
        field0Ptr_
     && timeIndex_ != this->time().timeIndex()
     && !(
            this->name().size() > 2
         && this->name()(this->name().size()-2, 2) == "_0"
         )
    )
    {
        storeOldTime();
    }

    // Correct time index
    timeIndex_ = this->time().timeIndex();
}

But due to my propably limited knowlwdge of C++, I failed to find the places where the above function is called and where the "storeOldTime()" is defined. Could anybody know about this issue? Thank you in advance!

OFFO

[deepsterblue]

It happens on-demand, whenever you invoke 'field.oldTime()'

If you look at the 'GeometricField:ldTime()' member function, it checks for the existence of field0Ptr_, and creates one if it doesn't exist. From that point on, it is updated on access only if the time-step has changed (by checking the value of timeIndex_, which is incremented every time-step)

[openfoammaofnepo]

Thank you, deepsterblue. I have not found a place where field.oldTime() is invoked. Could you please give me an example in it? Thank you very much. This is also my question so that I cannot understand when the old fields are stored in running OpenFOAM solvers.

OFFO

Originally Posted by deepsterblue

It happens on-demand, whenever you invoke 'field.oldTime()'

If you look at the 'GeometricField:ldTime()' member function, it checks for the existence of field0Ptr_, and creates one if it doesn't exist. From that point on, it is updated on access only if the time-step has changed (by checking the value of timeIndex_, which is incremented every time-step)

[deepsterblue]

You gave the example yourself - in your original post. When you have:

rho.oldTime() * vf.oldTime()

the storage of old-times is invoked at that point.

[openfoammaofnepo]

Thank you. Since the following is from the Euler time scheme:

rDeltaT*(rho*vf - rho.oldTime()*vf.oldTime())

So does it mean that when the time discretization is perdormed, the old fields will be stored? I originally thought that rho.oldTime() only extract the old rho fields fore previously stored fields.

Originally Posted by deepsterblue

You gave the example yourself - in your original post. When you have:

rho.oldTime() * vf.oldTime()

the storage of old-times is invoked at that point.

[deepsterblue]

Originally Posted by openfoammaofnepo

So does it mean that when the time discretization is perdormed, the old fields will be stored?

Yes - makes sense, doesn't it? If you don't discretize in time, then you don't need an old time, and that never needs to be stored. Simple and elegant.

[openfoammaofnepo]

It seems make sense. In one Solver, we will have several time discretizations, like in UEqn, rhoEqn, pEqn, for rho and the relavant quantities. If I follow your line of reasoning, the oldTime() will be store many times?

Originally Posted by deepsterblue

Yes - makes sense, doesn't it? If you don't discretize in time, then you don't need an old time, and that never needs to be stored. Simple and elegant.