[cojua8]

Hello, I'm trying to add a new member to KinematicParcel class of type List<scalar> to store all (or a defined number) the previous dt.

for this, I added, in file KinematicParcel.H:

...
protected:
            //- dt storage [s]
            List<scalar> dtList_;

public:
            //- Return const access to the dtList
            inline List<scalar> dtList() const;

            //- Return access to the dtList
            inline List<scalar>& dtList();
...

In file KinematicParcelI.H:

...
template<class ParcelType>
inline Foam::List<Foam::scalar> Foam::KinematicParcel<ParcelType>::dtList() const
{
    return dtList_;
}

template<class ParcelType>
inline Foam::List<Foam::scalar>& Foam::KinematicParcel<ParcelType>::dtList()
{
    return dtList_;
}
...

the in file file KinematicParcel.C (move function), I need to do this:

...
        if (p.onBoundaryFace() && td.keepParticle)
        {
            if (isA<processorPolyPatch>(pbMesh[p.patch()]))
            {
                td.switchProcessor = true;
                std::cout << "cambio de procesador" << std::endl;
            }
        }
        
        p.age() += dt;
        p.dtList().append(dt); 

        td.cloud().functions().postMove(p, celli, dt, start, td.keepParticle);
    }

    return td.keepParticle;
...

This works fine when I don't run my problem in multiple processors, but when running it in parallel, after going to a new processor it crashes.


How should I include dtList in the constructors so that it works as I expect and doesn't reset and crash when running in parallel?

Also, how can I set a fixed sized based on variable NB (shown on files attached)?

thank you!