[CGuaq]

Dear foamers,

I need to save different values during a time step to use them for the next time step, to do that I thought of using 2 volVectorFields defined in the constructor to "stored" values in vectors components:

sigma_n
(
	IOobject
	(
		"sigma_n",
		this->owner_.mesh().time().timeName(),
		this->owner_.mesh(),
		IOobject::NO_READ,
		IOobject::AUTO_WRITE
	),
	this->owner_.mesh(),
	dimensionedVector("sigma_n", dimless, vector(10e-12, 10e-12, 10e-12))
 ),

 u_n
 (
	IOobject
	(
		"u_n",
		this->owner_.mesh().time().timeName(),
		this->owner_.mesh(),
		IOobject::NO_READ,
		IOobject::AUTO_WRITE
	),
	this->owner_.mesh(),
	dimensionedVector("u_n", dimless, vector(0, 0, 0))
    )

Now my problem is to modified this volVectorFields in each cells and for each time steps. I use the following lines to access to values in cells in volVectorFields:

const volVectorField& sigma_n = obr.lookupObject<volVectorField>("sigma_n");
volVectorField& sigma_nMod = const_cast<volVectorField&>(sigma_n);
    
const volVectorField& u_n = obr.lookupObject<volVectorField>("u_n");
volVectorField& u_nMod = const_cast<volVectorField&>(u_n);

And futher I modify the values in each cells with:

u_nMod.primitiveFieldRef()[celli].component(0) = UTurb.component(0);
u_nMod.primitiveFieldRef()[celli].component(1) = UTurb.component(1);
u_nMod.primitiveFieldRef()[celli].component(2) = UTurb.component(2);
            	
sigma_nMod.primitiveFieldRef()[celli].component(0) = sigma_x;
sigma_nMod.primitiveFieldRef()[celli].component(1) = sigma_y;
sigma_nMod.primitiveFieldRef()[celli].component(2) = sigma_z;

I can compile my code and lauch simulation without errors but it doesn't seem to work ...
So I wonder if it's possible to save datas in volVectorFIeld between two consecutive time step with this method, if not can you help me to find another solution ?
I'am beginner in programming under OpenFOAM so any idea could be helpful for me !

Thanks in advance!

Best regards

[rku]

Hi Charles,

I am struggling to do a similar thing where I need the values of a vectorField to be carried over to the next timestep. Did you manage to solve your problem?

Thanks!

[CGuaq]

Originally Posted by rku

Hi Charles,

I am struggling to do a similar thing where I need the values of a vectorField to be carried over to the next timestep. Did you manage to solve your problem?

Thanks!

Hello rku,
To solve this problem I modified other files in the code (for turbulent dispersion) to add the vector fields I need as "input variables" in the turbulent dispersion model. With this solution this fields are automatically saved at each time steps and you can use them. But the different lines of code I added in my previous message can also do it. If you need more informations don't hesitate.

[rku]

Thanks for your quick reply!

I think I'll need to do some re-coding of other files as well.. The issue is the field I wanted to modify is vectorField which makes it more difficult in some ways. I'm trying now to generate a volVectorField which atleast allows me to use the oldTimes method. Lets see. Thanks anyways!