Page 58 - Fundamentals of Computational Geoscience Numerical Methods and Algorithms
P. 58

44              3  Algorithm for Simulating Coupled Problems in Hydrothermal Systems

                                                         3

                            4



                                               A
                          y



                                    1                            2



                       0                    x
            Fig. 3.2 Point A in a fournode quadrilateral element

                  →  →   →     →                                    →   →   →
            where 12, 23, 34 and 41 are vectors of four sides of the element; 1A, 2A, 3A
                →                                                        →
            and 4A are vectors of each node of the element to point A respectively; k is a
            normal vector of the plane where the element is located; λ i (i = 1, 2, 3, 4) are four
            different constants. Note that the left-hand sides of Eqs. (3.22), (3.23), (3.24) and
            (3.25) represent the cross products of two vectors.
              By using the global coordinates of point A and four nodal points of the element,
            Eqs (3.22), (3.23), (3.24), (3.25) and (3.26) can be expressed in the following form:


                        (x 2 − x 1 )(y A − y 1 ) − (x A − x 1 )(y 2 − y 1 ) = λ 1 ≥ 0,  (3.27)
                        (x 3 − x 2 )(y A − y 2 ) − (x A − x 2 )(y 3 − y 2 ) = λ 2 ≥ 0,  (3.28)

                        (x 4 − x 3 )(y A − y 3 ) − (x A − x 3 )(y 4 − y 3 ) = λ 3 ≥ 0,  (3.29)

                        (x 1 − x 4 )(y A − y 4 ) − (x A − x 4 )(y 1 − y 4 ) = λ 4 ≥ 0,  (3.30)

            where x i and y i are the global coordinates of nodal points of the element; x A and y A
            are the global coordinates of point A.
              Note that when point A is located on the side of the element, two vectors related
            to a node of the element are coincident. Consequently, the corresponding λ i to this
            node must be equal to zero. Since Eqs. (3.27), (3.28), (3.29) and (3.30) are only
            dependent on the global coordinates of five known points (i.e., point A and four
            nodes of the element), they can be straightforwardly used to predict the element, in
            which point A is located.
              It must be pointed out that, using the point-searching strategy expressed by
            Eqs. (3.27), (3.28), (3.29) and (3.30), it is very easy and accurate to relate a point
            in mesh 2 to an element (containing this point) in mesh 1, since Eqs. (3.27), (3.28),
            (3.29) and (3.30) only involve certain simple algebraic calculations which can be
            carried out by computers at a very fast speed. Thus, to translate/transfer the solution
   53   54   55   56   57   58   59   60   61   62   63