[A_Pete]

Hey everyone,

I am right now trying to understand the hitWallFaces() function in particleTemplates.C of the lagrangian tree. It is checked whether a particle hits a face that belongs to a wall or not. I think that I understand most of the code until the tetLambda() function is called as follows:

scalar tetClambda = tetLambda
(
    tetIs.tet(mesh_).centre(),
    toPlusRNHat,
    0,
    n,
    f[tetIs.faceBasePt()],
    cellI_,
    fI,
    tetIs.tetPt(),
    lambdaDistanceTolerance
);

Where IMO tetIs.tet(mesh_).centre() is the centre of the considered tetrahedron and

toPlusRNHat = to + r*nHat

is defined before. "to" is the position the particle is supposed to move to after the time step, "r" is the particle radius and "nHat" is the normalized normal vector of the considered triangle.

What tetLambda now does (as far as I've understood and please correct me if I'm wrong) is to check whether "toPlusRNHat" is behind the considered wall face outside of the domain. I have checked for multiple triangles and I see that the triangle normal vector "nHat" is always pointing into the simulation domain. To me this means that "toPlusRNHat" is always more inside of the domain than "to" (the wanted position of the particle when there is no hit of a face) is. To my understanding it would make more sense if

toPlusRNHat = to - r*nHat

(see the minus sign in front of the second term instead of a plus sign).

Since then the particle would mean to hit a face as soon as its outer radius hits the face.

Does anybody have any expierence with this class? Thank you very much in advance.

Best regards,
Andreas

[A_Pete]

Has anybody worked on this and some hints for me?

In my opinion, since the normal vector "n" of a triangle points inwards into the domain (this may be contrary to face normal vectors), toPlusRNHat should be defined as:

toPlusRNHat =to - r*nHat

So that the actual position the particle wants to move to is shoved towards the face by one radius in normal direction.